3PAR, Inc.

Defining a New Tier of Cloud Storage: The Data Center Tier

2010-03-15T10:00:00Z

Cloud storage is generally broken out into two broad classifications: public storage clouds and private storage clouds. But as cloud storage technologies continue to mature, it is quickly becoming evident that these two categorizations are insufficient. Recent offerings from providers like 3PAR, EMC and IBM demonstrate that a new cloud storage tier is emerging. This tier extends the benefits of cloud storage up the application stack so that data center applications can also take advantage of the cloud storage architecture.

The concept of cloud storage is resonating with all size organizations for a number of reasons but possibly the biggest is that it makes managing data much easier. By creating a single logical pool that stores and manages data, the effort associated with routine, ongoing maintenance tasks such as data migrations, system upgrades, and even determining on which storage device to store the data is markedly reduced or even goes away.

However most cloud storage (public or private) solutions are designed the way they are for one simple reason: they are designed to host archival, backup and infrequently accessed data. Since 80% or more of an organization's data fall into one or all of these three categories, current cloud storage solutions primarily seek to deliver acceptable performance, high capacity and low cost. To do this, first generation cloud storage offerings use inexpensive servers, storage and network interfaces to meet these demands.

These requirements change as you go up a notch and start to apply cloud storage architectures to data center environments. Data center managers also want to harness the benefits of cloud storage (simplified data migrations, system upgrades and a single logical pool of storage) but need other characteristics that typical cloud storage solutions do not provide. New characteristics that define a data center cloud storage offering include:

Block based interfaces. While some private storage cloud offerings offer iSCSI as a block based interface, this protocol is generally viewed as insufficient for performance intensive applications. So any cloud based offering that is to be used by a data center will need to support protocols such as Fibre Channel (FC) or FC over Ethernet (FCoE) to satisfy its performance demands.
Multiple tiers of storage. First generation cloud storage offerings primarily use SATA drives to store data because they are high capacity and low cost. The downside is that these drives are relatively low performing. While data center managers certainly desire to store some of their inactive production data on this tier of storage, they also have a percentage of their application data that requires higher performing FC, SAS and even new SSD. So whatever cloud storage solution they deploy has to support all of these tiers of storage capacity.
Dynamic placement of application data on the appropriate tier of disk at the right time. Just having multiple tiers of storage capacity is not enough. There must be intelligence added into the storage system that automatically places data on the appropriate tier of storage according to the application's needs. This placement needs to be policy driven, done automatically and adjust dynamically to changing application requirements.
Simplified storage provisioning in block-based environments. This is one of largest hurdles that this emerging data center cloud storage tier has to overcome. The need for this feature is driven by server virtualization where a single physical server may host multiple virtual machines each with its own unique identifier. When this host is connected to a storage cloud, performing tasks such as zoning on FC switches and directors and applying LUN masking to each LUN become numerous and complex. While it can be done manually, it becomes very time consuming and risky as it introduces the potential for human error. To avoid this, the data center tier must provide a means to simplify and automate this very cumbersome and time consuming process.
Risk-averse. This may sound like an oxymoron in the data center world - how can you implement a cloud storage solution and minimize risk while doing so? However that is exactly what is required in order for these managers to feel comfortable adopting a data center cloud storage solution.

The new problem that results from this data center tier of cloud storage is that in addition to only one logical pool of data being created, it also creates a storage system that functions logically as one entity. This means tasks that are currently segregated (firmware updates, zoning and LUN masking changes, etc. because they are on physically different storage systems) are now consolidated onto one unit.

This gives rise to the potential that a single mistake or even a series of unrelated mistakes can bring down the entire storage cloud - an unacceptable proposition in data centers. So in order for a cloud storage solution to achieve the "Data Center Tier" classification, it must provide checks and balances such that data center managers have the flexibility and ability to control what changes are implemented, when they are implemented and under what circumstances.

The concept of cloud storage has infused new enthusiasm into a part of the technology industry that is historically perceived as a bit more conservative. But even now that conservative nature is on full display as cloud storage starts to move up the application stack into enterprise data centers. Notice that it is not unknown providers that are looking to deliver a data center tier of cloud storage but recognizable names and emerging brands within these established storage providers such as the 3PAR InServ Storage Server, the EMC V-Max, and the IBM XIV.

This is where data center managers need to be on their guard. As this new "Data Center Tier" of cloud storage continues to mature, each provider is developing its own set of options to solve these aforementioned challenges. As they do, what will separate the best products from the also-rans are those that provide the right underlying framework (hardware and software) that make this new data center tier of cloud storage affordable to obtain, easy to implement and manage, and does so without putting their enterprise at risk.

3PAR's New Adaptive Optimization is like Butter

2010-03-08T13:00:00Z

The initial hype around solid state drives (SSDs) is starting to settle down. The performance benefits, costs, "gotcha's" and use cases of SSDs are now better documented which is resulting in new implementations of SSDs that emphasize their benefits while mitigating their drawbacks. A prime example of this is today's announcement from 3PAR that adds support for SSDs to its InServ Storage Servers but more importantly provides users with a means to optimally and simply take advantage of them.

A recent trend (the last 3 - 6 months) in the implementation of SSDs is a move away from exclusively deploying SSDs as a permanent repository for application data. The reason behind this is simple. The majority of application data (90% to as high as 99%) does not require the performance of SSDs nor can businesses absorb the high cost associated with putting all of this data on SSDs.

This is leading to SSDs being deployed and managed as a cache within storage systems. Utilizing SSDs in this fashion only the most active application data is stored on SSD while less frequently accessed data is stored on more economical SATA, SAS or FC hard disk drives (HDDs). (I refer to FC and SAS as "more economical" in the context of being compared to SSD.)

This approach sounds appealing but the problems associated with deploying SSDs solely as a cache without any corresponding intelligence that controls data placement are numerous. These challenges include:

Identifying and moving the most active application data from current HDDs to the SSD.
Moving this data back off of the SSD to the HDDs as the data becomes inactive.
Scaling the SSD solution up as requirements change.
Automating this data movement so it occurs without human intervention
Providing overrides and policies so that some data movement is subject to human control.

In other words, implementing SSDs as a cache is fairly easy; implementing SSDs so applications can effectively use them, costs are kept down, and performance is expedited for the widest range of applications is more difficult.

It is these multiple concerns that 3PAR addresses with today's announcement of the addition of SSDs and adaptive optimization feature to its InServ Storage Servers. Of the two announcements, it is 3PAR's adaptive optimization feature that merits the most attention from end-users.

Adaptive optimization is to SSDs what butter is to lobster. You can certainly eat lobster without butter but butter is what gives lobster its succulent taste. Those deploying SSDs in a 3PAR system without adaptive optimization will have a similar experience - you cannot realize the full benefits of SSD.

Adaptive optimization builds upon 3PAR's existing chunklets, dynamic optimization, regions and wide striping technologies.

Chunklets are little, virtual pieces of physical disk.
Regions are a narrow stripe of capacity across many chunklets and are assigned to logical disks that are used to create volumes.
Wide striping takes advantage of these underlying "chunklets" and "regions" by spreading them across multiple disk drives in the 3PAR storage server to improve overall application performance even when lower performing disk drives are used.
Dynamic optimization then moves regions associated with an entire volume to an appropriate tier of disk based upon their applications' access patterns and usage.

Adaptive optimization builds upon these four technologies. Adaptive optimization monitors the I/O rates and activities of each specific region. The real advancement that adaptive optimization delivers is that it manages regions independently instead of as a group associated with a single volume.

This advancement enables 3PAR to place move data associated with a specific region on any tier of disk (SSD, FC or SATA) to a different tier according to that region's performance characteristics. This creates the following new possibilities:

Availability of SSD to all application data on 3PAR volumes. Because all 3PAR volumes are made up of regions, as soon as SSDs are placed in a 3PAR system and the adaptive optimization is turned on, all 3PAR volumes can be associated with the Adaptive Optimization profile and have access to the SSD drives.

Selective allocation. Just because SSDs are installed on the InServ Storage Server and adaptive optimization is turned on does not mean a volume gets unfettered access to the SSDs. Administrators can set up policies that control when and how much application data on a volume can reside on the SSD and they can even prohibit certain volumes from storing any data on the SSDs at all.

Controlled data movement. Using SSDs as a cache sounds appealing until one considers all of the data movement that needs to occur between these different tiers of disk and the overhead that this can introduce on the storage system.

To minimize this, adaptive optimization supports the creation of profiles that consist of specific tiers which have policies associated with them. These policies control when data movement occurs between tiers for specific volumes such that if a region of data that normally experiences minimal activity suddenly gets busy, that region does not immediately get placed on SSD. Rather it stays on SATA or FC for a defined period of time until it passes a certain threshold.

Conversely, if an extremely active region that resides on SSD goes abnormally dormant for a period of time, 3PAR does not automatically migrate that data to SATA disk. Instead it again leaves the data there until it crosses a specified threshold before it vacates it from the SSD disk.

3PAR has also introduced a feature called QoS Gradients which allows an administrator to bias volumes toward performance or cost resources, depending upon the application service level objectives. A performance gradient would be used for an application with a high SLA or a cyclical application that demands rapid response to a spike in activity. In contrast a cost gradient would be used where cost is the primary objective and an application with have a SATA-stickiness associated with it.

Reduced total cost of ownership. Of course, maybe the message that will resonate most with end-users is that adaptive optimization actually lowers storage costs.

Using adaptive optimization, 3PAR can recognize the most active regions of data within volumes and move them to SSD while placing inactive regions of data on lower cost SATA disk. This flexibility gives organizations new found flexibility to start to consider SSD-FC-SATA storage system configurations that look something like 5% SSD, 15% FC and 80% SATA even for their production application workloads.

This combination should serve to minimize the amount of SSD needed, decrease the number of FC drives needed, and increase the usage of SATA drives that will ultimately serve to drive down storage costs by 20% or more.

One final item of note regarding the introduction of SSDs in the 3PAR InServ Storage Server: 3PAR is able to use STEC's new Mach 8 IOPS 50 GB drive in lieu of its Zeus IOPS drives. While these Mach 8 drives are about 1/3 of the capacity of 147 GB Zeus IOPS drives, they are also 1/10th the price so organizations can add SSD on a more granular basis to the 3PAR system at a lower price point.

Today's announcement of 3PAR's support for SSD on its InServ Storage Servers and its complementary Adaptive Optimization feature is a perfect illustration of how the hype surrounding SSD is coming to an end and the reality associated with implementing SSDs is coming into focus. Companies want SSD performance for their applications but they want it without the huge financial outlays.

Right now SSDs are sizzling but adaptive optimization is what make SSDs palatable to the corporate taste buds. 3PAR's combination of adaptive optimization and SSDs give users get the performance they want, the cost savings they need and the simplicity of implementation and management that continue to set the 3PAR InServ Storage Servers apart from others in this tier of storage.

If Given a Choice - Choose Array-based Thin Provisioning over VMware's Thin Provisioning

2010-02-25T11:00:00Z

Over the last few years thin provisioning has steadily moved into the main stream of storage management - so much so that not only has it found its way onto many leading storage systems but into operating systems as well. Clearly one of the largest endorsements of using thin provisioning at the operating system level came last year when VMware announced its inclusion of thin provisioning as an option within vSphere 4.

But with thin provisioning now available at both the OS and storage system levels, organizations need to quantify at which of these two levels that they can derive the greatest number of benefits from thin provisioning or if, in fact, there are sufficient benefits at both levels to implement both storage and VMware thin provisioning .

A place that is as good as any to start in trying to understand the benefits for each is a video that was embedded in a blog on the StorageRap website on September 1, 2009. This video contrasts the benefits of using VMware's thin provisioning with using the thin provisioning feature found on the 3PAR InServ Storage Server.

The initial scenario that the presenter, 3PAR's Michael Haag, illustrates is when the storage allocated to VMware is a fat volume (i.e. NOT thinly provisioned) from traditional storage. All of this storage capacity is discovered and allocated by the VMware file system (VMFS) and brought under its management.

It is in this scenario that the primary value of VMware's new thin provisioning feature comes into play. As new virtual machines and their associated VMDK files are created, VMFS will only allocate as much physical space as each individual VMDK file actually needs, reserving only enough space for the VMDK's data. This results in storage savings and enables vSphere to more efficiently meet the storage requirements of the virtual machines it hosts as well as open the door to hosting more VMs.

However there is still a very real up front storage cost with VMware's thin provisioning implementation: all VMFS capacity must be provisioned (physically) up front instead of as it is needed for each VM and VMDK.

So while VMware vSphere's thin provisioning feature certainly helps organizations get more value from "fat" storage with a thin methodology from VMFS to the VM, it still leaves the door open to increase storage utilization with array-based thin provisioning. Array-based thin provisioning not only associates physical capacity with the VMDK only as writes occur, but does so with VMFS as well so there is no up front allocation or waste.

Consider this scenario that occurs every day in organizations of all sizes. It is only natural for an application administrator to request more storage capacity than the application needs. In this case, VMware's thin provisioning helps alleviate this over provisioning request by enabling storage administrators to thinly provision VMFS' allocated storage capacity to the VM-based application as is written.

However the storage administrator is still forced to hedge his bets. While he may suspect the application administer has over estimated how much storage capacity he needs, the storage administrator needs to over provision the networked storage assigned to that vSphere server to account for the situations where the application owner's estimates are correct or even too low.

It is for this reason that the presenter in this video as well as other virtualization experts not affiliated with 3PAR recommend, at a minimum, using thin provisioning at the array level. VMware thin provisioning can be used in conjunction with array-based thin provisioning; however, thin provisioning at the array level should be used at minimum because it provides the most efficient use of storage capacity for three reasons:

Storage capacity is only allocated and consumed (to VMFS and the VMDK) when data is actually written by the application on the VM to the attached storage system.
Storage capacity does not need to be over-provisioned
Storage management can remain a function of the storage team for those enterprise organizations that have this separation of duties. However because of 3PAR's integration with VMware's vCenter, organizations can also monitor, manage and report on storage through this management console.

Finally there are cases where it makes sense to use the two thin provisioning solutions together. The additional overhead of adding VMware TP on top of array-based TP can be beneficial in cases where separate Server and Storage Administrators want the flexibility to over-provision at their respective layer.

The growing acceptance of thin provisioning in the data center is leading to increased cost savings and efficiencies but, as this example with VMware vSphere illustrates, sometimes it is not always clear which is the best way to implement it.

The short answer is that if the storage system you are using now does not offer thin provisioning, using vSphere's thin provisioning feature is clearly the way to go. But in circumstances where you have both VMware and a storage system from 3PAR available to you and if thin provisioning can only be used in one place, then it should absolutely be done at the array-based layer to maximize utilization and minimize upfront costs.

Administrators can then decide to deploy VMware thin provisioning on top of 3PAR thin provisioning if it provides enough management benefits to the VMware administrator to outweigh the need to manage thin provisioning in two places.

A Capacity Savings Guarantee that Delivers like the Energizer Bunny

2010-02-18T14:00:00Z

Capacity savings guarantees are still a relative rarity but can be more easily found as technologies like deduplication and thin provisioning find their way onto primary storage systems. However the bloom can quickly come off the rose when one starts to dig into the details associated with these guarantees. That's why 3PAR's recent 50% capacity savings guarantee announcement stands in stark contrast to other similar guarantees as its savings are not just a one-time event but potentially keep going and going and going.

The number of storage providers guaranteeing any sort of capacity savings is still few and far between. Only three other storage providers besides 3PAR currently make this type of guarantee with the requirements for meeting the conditions of each of their guarantees varying widely. For example:

One provider's guarantee is only applicable when its storage system is used in conjunction with a server virtualization operating system from Citrix, Microsoft and VMware. Further, when one reads this provider's Technical Guide associated with its guarantee, it goes on to specifically EXCLUDE large database and Exchange deployments from this guarantee. In addition, professional services are a required part of the guarantee program.
Another storage provider promises new customers a 40% reduction in storage capacity versus their current configuration. This guarantee is notable in that it makes this promise to customers even if they are already using thinly provisioned volumes. The catch with this guarantee is that it is only measured at the time when their storage system is installed and data is migrated to it.
The third one guarantees a minimum 80% utilization rate when their storage system is used with no performance degradation. Again, a notable guarantee but one of its caveats is that the customer must notify the storage provider if the customer makes any changes to the environment that may affect the storage system. My own experience tells me that this guarantee is not practical to enforce since often the individuals responsible for the storage are the last ones to hear about changes to applications using the storage system.

Now to give credit where credit is due, at least one of these three storage system providers is willing to put some skin in the game with an actual contract accompanying their guarantee. The others are guarantees in the 'marketing' sense but have no legal foundation. Most of the time when users are looking to implement a new storage system, they begin with the assumption that they will not reclaim any storage capacity and only expect to add more.

While some may argue that any guarantee is better than nothing at all, clearly the conditions of these guarantees exclude the situations that many end-users commonly face. In cases where users can take advantage of them, the practical implementation of these guarantees still leaves something to be desired.

This is really what distinguishes 3PAR's 50% capacity savings guarantee from the other guarantees already discussed: the practicality of its implementation and ongoing management. This is found in three ways:

First, 3PAR customers will see at least a 50% capacity savings when converting from existing "fat" volumes on traditional storage arrays to thin volumes on any 3PAR InServ Storage Server. What is important to note is that 3PAR places very few conditions around its guarantee as 3PAR extends this guarantee to all types of application servers whether or not they are virtualized. No professional services are required, and 3PAR is willing to stand behind its program with a contract versus simply a marketing statement.
Second, 3PAR makes it possible for this capacity savings not to be a one-time event. One of the reasons thinly provisioned volumes get "fat" over time is because storage capacity that is used and then freed up by the application is never reclaimed by the storage system. 3PAR helps to eliminate that problem with its Thin Engine^TM technology that dedupes "zeros" associated with deleted data on a volume. In this way, organizations will continue to realize capacity savings over time, not just when data migrations are performed.
Third, there is no need to call 3PAR every time changes are made to the environment to ensure compliance. While 3PAR's Thin Engine^TM technology automates the recapture of freed space, its Thin Built In^TM technology distributes the application data across all available drives in its system. This technique ensures optimal performance for each application without requiring a professional services engagement every time a change to the system is made.

3PAR's guarantee that users will realize a 50% reduction in storage capacity by migrating their application data to its InServ® Storage Servers is certain to draw in new customers. However what new customers should find encouraging and 3PAR's existing customers already know are that 3PAR's savings are not limited to certain application servers, one-time data migration events or require ongoing professional services to implement and manage.

Rather the technologies found in 3PAR's InServ Storage Servers deliver the one-time capacity savings that users need today to help justify 3PAR's adoption while keeping storage growth under control in the future that will continue to justify their decision to choose 3PAR both now and going forward.

3PAR Introduces a Dose of Reality into Virtual Data Center Management

2010-01-26T11:00:00Z

The benefits of virtualization - server or storage - start to loose some of their shine as organizations come to grips with the reality of actually managing a virtualized environment. As the move from implementation to management, the hidden issues of managing virtualized environments become clearly exposed. It is for these reasons that storage providers like 3PAR are more tightly integrating their InServ Storage Servers with VMware's vCenter Server management console to provide better management of virtual data centers.

Transitioning to a virtualized environment creates new requirements for the effective management of the underlying storage resources that are virtualized. This can become a challenge as many storage management tools were built for physical environments which make them of only marginal assistance when managing a virtualized environment.

Storage and system administrators are then left with an incomplete view into how physical resources in the virtual environment are used which creates knowledge and visibility gaps. These gaps result in:

Increases in the amount of time and effort required to manage the environment
The inability to definitively associate storage resources with specific virtual machines (VMs)
The creation of complex, error-prone manual processes that try to do these virtual-to-physical mappings

Developed in part to help eliminate these types of issues, VMware vCenter Server provides an open plug-in architecture that supports the addition of new functionality from VMware and its partners. Through vCenter Server APIs, VMware's partners can give VMware administrators new ability to manage their virtual and physical infrastructure through the vCenter console.

It is integration with these vCenter APIs that 3PAR announced yesterday. 3PAR's new plug-in for vCenter Server provides an integrated view of individual VMs and the storage resources associated with them. In so doing, it bridges this visibility gap that exists in virtualized environments and provides a complete end-to-end mapping that graphically illustrates exactly what storage resources are associated with individual VMs.

Since all of 3PAR's T-Class and F-Class storage servers support thin provisioning, 3PAR's vCenter plug-in enables administrators to, at-a-glance, determine whether a LUN assigned to a VM is 'Fat' or 'Thin'. Then if it is a "Thin' LUN, VMware administrators can see what storage resources a given VM or VMDK is actually consuming as well as document the storage savings that they are achieving, an excellent benefit for those responsible for charge back of a "slice" of the virtualized environment back to users.

3PAR's integration with vCenter also opens the door for 3PAR to introduce new management functionality which it promptly did in the form of VMware Recovery Manager. While 3PAR has supported snapshot functionality on its InServ storage servers for some time, storage administrators had to set up and schedule snapshots to occur from the 3PAR management interface. By integrating with vCenter, 3PAR extends these snapshot setup and scheduling capabilities so they can be performed by VMware administrators within vCenter.

The benefits of the 3PAR-vCenter integration go beyond just providing a convenient central management interface for administrators. By using vCenter as the platform to configure these snapshots, VMware administrators can also take consistent snapshots of the VM. Since vCenter is scheduling the snapshots on the 3PAR storage server, it first pauses the application on the VM so that the snapshot created is a restorable image of the application on the VM.

This integration reduces recovery management within VMware to almost a point-and-click operation. All VMware administrators essentially now need to do within vCenter is select the VM they wish to recover and they can essentially bring that VM back online rapidly. While there are a couple of other steps that administrators have to do once that step is done, these all can be done through vCenter.

If anything, possibly the biggest challenge that organizations may face is how to best internally administer this functionality. Often there is a dividing line, spoken or unspoken, between what tasks server and storage administrators perform. By making all of this functionality available from the vCenter console such that a single administrator can perform them, organizations need to give some thought to how they best want it administered. But probably better to have this choice than not having this functionality available at all.

The creation of virtual data centers is a new reality that many organizations are now in the process of putting in place. But once they are in place, the new reality of managing these virtualized environments immediately confronts them.

3PAR's integration with vCenter and its new Recovery Manager feature provides a dose of reality in today's world that many administrators are seeking out. vCenter already gives them the foundational platform to manage their virtual environments through a single console. Now with 3PAR's advanced integration with vCenter, they get a comprehensive view into both physical and virtual infrastructures, an end-to-end mapping of their virtualized environment, and a new tool that enables them to schedule the snapshots and provide rapid recoveries of individual VMs on their own.

3PAR Succeeds because it Lightens the Load without Needing to Darken the Skies

2010-01-13T11:00:00Z

Good times or bad, all size organizations need data storage. The difference is that when times get tough as they were in 2009, we start to see those storage providers that deliver a good value for the dollar get the nod over those that deliver only so-so value. But what is notable about 3PAR, which recently received some positive comments from Barron's, is that 3PAR is making inroads in the high end of the storage market where start-ups are not supposed to succeed.

In the last week of December 2009, Barron's commented that 3PAR has "cheaper, more advanced technologies that larger rivals initially ignored", its data storage systems "do more with less", they are "unusually flexible" and support a "multitenant environment".
Given the Barron's article, I wanted to examine further how 3PAR's technologies are contributing to the company's appeal.

3PAR has managed to accomplish something that few other providers of storage systems have succeeded in doing: break into high end data center environments as a start-up without first starting in the small and midsize space. The article mentions that 3PAR has picked up some impressive enterprise customers such as MySpace, Verizon, Credit Suisse and Priceline.com but it fails to mention or examine just how rare it is for any start-up storage provider to experience any type of success in the enterprise data center space.

Talk all you want about "cheaper, more advanced technologies", "doing more with less", and being "unusually flexible" but to succeed in enterprise environments, it takes more than that. In these data centers, storage providers need dependable and reliable enterprise solutions which traditionally has been a prerequisite to success.

3PAR has succeeded in a rather unique way by delivering dependable and reliable enterprise solutions without the need for professional services. Some of its competitors continue to get enterprise business with what I consider inferior products because they make guarantees to enterprise IT management and staff that they will "darken the skies" with professional services people should they ever have a problem with their product.

While this sales tactic sounds good in theory, there are only two problems with it. First, to make this offer means that they (the storage provider) are assuming that you (Mr. Customer) will at some point have a substantial problem with their storage system or that the solution presented by sales somehow will fall short of its intended objective. This also makes it a necessity that they put such a guarantee on the table so that when this unforeseen and unexpected disaster strikes, the customer has assurances the storage provider will be there to make things right.

But from where I sit, this guarantee is somewhat disconcerting. If you are buying the right product for your environment, one would think the need to "darken the skies" with professional services help would be an isolated and rare incident, not a regular occurrence.

Having talked to both 3PAR and a number of 3PAR storage customers, such incidents appear rare. Richard Siemers, a storage admin with Pier 1 Imports, probably puts it best in regards to 3PAR's services model. He says, "Professional services (PS) are something I prefer to do without. A 'Statement of Work' is not intended to be a customer friendly document. Often the 'professional' sent is nothing more than a below average tech with a collection of good instructions and a speed dial to a real professional. I appreciate 3PAR for the way they operate without PS, and I appreciate the SE we have assigned to our account."

Second, many of the storage systems sold into today's enterprise data centers were initially designed to work in mainframe environments and then retrofitted to work with today's distributed system environments. This is not necessarily wrong but a certain amount of baggage accompanies such design decisions. This is largely reflected in the level of complexity and amount of knowledge that one must first possess in order to setup and manage these storage systems.

In the past, this may have been less of an issue since many of the individuals responsible for managing these storage systems had years of expertise and technical ability that they acquired while managing storage on the mainframe side of the house.

But that experience does not come easily, data stores are burgeoning and organizations need storage solutions that match their virtualization requirements. Administrators no longer have the time and patience to learn everything there is to know about storage and the speed of today's business world is forcing them to select storage solutions that are designed - not retrofitted - to meet today's requirements.

It is these demands that 3PAR's InServ Storage Servers meet. They are easier to manage, more space efficient, handle multiple application workloads, require less time to operate and are a help rather than a hindrance to the organization. In short, they better match the skill sets and time constraints of the administrators tasked with the responsibility of managing them.

From a technology and support perspective, 3PAR is succeeding where many others have tried and failed. And while some speculate that 3PAR may be negatively impacted by its competitors as the economy begins to rebound in 2010, don't count on it.

3PAR delivers the high availability, reliability and support that enterprise environments demand but its storage servers are architected to deliver the new levels of flexibility and ease of use that today's enterprise storage administrators need. It is precisely because 3PAR delivers these features for today's more distributed and increasingly cloud-oriented enterprise data centers that it should continue to succeed against incumbent storage providers without the requirement for professional services engineers to darken enterprise data centers to keep everything up and working.

Tiered Storage Gets Smarter and Easier in One Fell Swoop

2009-12-07T13:00:00Z

Tiered storage is gaining momentum because of the ease in which storage systems make it possible for organizations to implement it. One way that storage system providers offer this feature is through the automated redistribution of volumes on an appropriate tier of disk according to preset policies. However since tiered storage's introduction, new customer requirements for managing it have emerged that dictate tiered storage software found on storage systems get smarter while at the same time making it easier for users to manage.

Using native storage system software to automate the redistribution of data on the appropriate tier of storage within them started in earnest this past decade. Driven primarily by the difficulty and labor involved with implementing tiered storage at the host level and the economic benefits that organizations can derive from using a mix of FC and SATA disk drives within a storage system, it only made sense for storage systems such as the 3PAR InServ Storage Server to offer tools that automate data mobility across disk tiers.

One method used to implement tiered storage is through the policy based tiering of data. Specific policies are assigned to storage system volumes so that data within that volume is always placed on a specific tier or tiers of disk to meet some combination of application availability and performance requirements. Administrators can then set policies so that these volumes are moved from one tier of disk to another tier in anticipation of current or future applications needs.

Normally, these types of data migrations are either not possible or extremely complex as they involve time consuming and often disruptive data migrations that move the volume from one tier of disk to another and then back again.

However, using a feature such as Dynamic Optimization that is found on 3PAR InServ Storage Servers, administrators can with single command move infrequently accessed data that is stored on SATA disk to higher performing FC disk in anticipation of end of quarter or end of year processing and then move it back to SATA disk once this processing is done.

They can also use it to non-disruptively accomplish other tasks such as data lifecycle management by changing the volume's underlying RAID levels, subsystem failure protection levels, drive types, and radial placement of data or increasing the stripe width across disks and controller resources, thereby aligning storage costs with the value of the data being served.

It is because of these types of capabilities that Dynamic Optimization has become one of the most popular tools on the 3PAR InServ Storage Server among its customers in the four (4) years since Dynamic Optimization's 2005 release. However during that span of time, many of 3PAR's customers have grown their 3PAR storage implementation which has created new opportunities for tiered storage functionality.

Specifically, as customers added more storage capacity in the form of disk drives to a storage system, they might also want to take advantage of the additional performance benefits that these new disk drives can deliver. The 3PAR array is architected in such a way that as resources are added to the system, the performance of all volumes improves.

To achieve this, the volume needs to span as many disk drives in the storage system as possible - existing and new. While accomplishing this is simply done with Dynamic Optimization, 3PAR saw an opportunity for policy management and multi-volume automation to further accelerate such tasks.

The availability of Policy Advisor for 3PAR's Dynamic Optimization feature provides for policies to be set around volume layout and can launch optimizations for hundreds of volumes with a single command, redistributing the data on each volume across both existing and new disks. Policy Advisor helps organizations in several important ways:

Users set policies by storage tier. Policy Advisor allows the user to set thresholds, above which volumes will be targeted for redistribution across newly available resources. This can be specified for each storage tier within the InServ.
It analyzes all of the volumes configured on the array and prioritizes how to proceed. Before Dynamic Optimization redistributes a volume, Policy Advisor analyzes the existing volumes. Once this analysis is complete, it generates a report that identifies which of the existing volumes, based on pre-set policies, would most benefit from a dynamic redistribution of data across the new and existing disk drives and in what order volume redistribution should occur.
Automate the optimization or customize when and on what volumes the optimization occurs. Once Policy Advisor completes its analysis, the user can launch in a single command the optimization of all identified volumes which would benefit from redistribution. Alternatively, organizations can schedule the optimization to occur in any of a number of alternative ways. For instance, they can schedule it to run during scheduled maintenance windows or to only optimize a subset of volumes that would most benefit from optimization.

Moving tiered storage software from the host servers to external storage systems earlier this decade resulted in the broader adoption of tiered storage software at the storage system level by organizations as a whole. While data mobility still limits tiered storage deployments on a variety of platforms, for those vendors making it simple to implement, non-disruptive and completely automated, user interest in this technology is growing significantly.

However data storage technology is a rapidly changing space and what was innovative a few years ago can often benefit from additional breakthroughs in virtualization and automation technology. Tiered storage software is no exception from this basic truth in technology.

The introduction of Policy Advisor from 3PAR for Dynamic Optimization on its InServ Storage Servers represents a significant step forward in the evolution of tiered storage software in the following way: in order for tiered storage software on storage systems to stay relevant in changing or high growth environments like the virtual data center, policy management and multi-volume automation are critical.

Policy Advisor enables 3PAR's Dynamic Optimization software and tier management to move to the next level on its InServ Storage Systems. Tiered storage software gets smarter, it remains easy for organizations to implement and use on an ongoing basis and it gives organizations the flexibility to grow without having to "manage" storage.

Autonomic Groups Feature Strikes a Blow to Server Virtualization Complexity

2009-11-10T11:00:00Z

High-profile case studies about companies such as Google and Amazon document how they are leveraging the cost reductions that clustered virtualized servers make possible to improve reliability, scalability, performance and capacity. But as the clustering of virtualized servers becomes more prevalent for these reasons, the overhead associated with storage administration in these environments only increases. Provisioning storage volumes to clustered virtualized servers calls for more upfront planning; it takes longer due to the complexity involved; and, the probability of human error becomes difficult to avoid.

Clustering virtualized servers for live migration, availability, failover, disaster recovery and the other reasons mentioned is becoming almost a must-have due to the number of virtual machines (VMs) that individual physical servers host. Yet what companies can fail to account for as they cluster these virtualized servers is the complexity that sharing networked storage volumes among these virtualized servers can create.

For example, suppose you have five (5) clustered virtualized servers that each need shared access to ten (10) storage volumes on a networked storage system. To share these 10 storage volumes among these 5 virtualized servers, you will need to issue commands granting each virtualized server the permission to access each of these storage volumes. In short, up to 50 provisioning actions requiring hundreds and hundreds of commands may need to be issued and done correctly for this configuration to work.

Further, having personally done tasks like this in the past, I am describing a best case scenario where everything goes well. In a worst case scenario where the wrong commands are issued, I have seen it take weeks or even months to complete.

This problematic nature of correctly assigning these different storage volumes to all of these clustered virtualization servers suggests that companies identify and implement more simplified mechanism for storage provisioning and ongoing management in these environments.

This simplification is what today's Autonomic Groups feature announcement from 3PAR provides by removing much of the complexity that is commonly associated with storage provisioning in virtual environments (e.g. VMware) and database environments (e.g. Oracle RAC).

Perhaps the largest benefit it provides is that it reduces the complexity associated with storage provisioning in clustered virtualized server environments down to three "clicks".

Here is what changes from before. Rather than trying to grant each virtualized server access to each storage volume, 3PAR users can in one click put all virtualized servers into a single Host group. Similarly, all of the individual storage volumes can now be created and put into a single Storage Volume group in another click.

Now when storage is assigned to these virtualized servers, rather than trying to do one-to-one mappings between each virtualized server host and individual storage volume, a single, final click in the GUI is all that needs to be issued to map the Host group to the Storage Volume group. Automation within 3PAR's operating system then presents this Storage Volume group to the Host group which preserves and presents the same LUN ID for this Storage Volume group to all of the servers in the Host group.

Autonomic Groups provide three other distinct benefits going forward as well:

When a new virtualized server host needs access to the Storage Volume group, administrators only need to add this new server to the Host group by issuing a single command. All of the storage volumes in the Storage Volume group get exported to the new server automatically.
Conversely, new storage volumes can be easily exported to all servers in the Host group by simply adding the new storage volume to the Storage Volume group. Again, this can be done with a single command and is automated, so no additional work is required on the part of the administrator.
Snapshots for data protection can be taken of the entire Storage Volume group in a single click with complete consistency across the group. 3PAR also announced Scheduler that automates the creation and deletion of Virtual Copy snapshots and can be used in conjunction with Autonomic Groups.

The Autonomic Groups feature still requires administrators to put specific hosts or storage volumes in their respective Host and Storage Volume groups during setup. However the time that this takes is literally seconds which is a fraction of what is normally required to assign one virtualized server its individual storage volumes under the old model.

The amount of risk that the Autonomic Groups feature removes coupled with the improvements in administrative efficiency that is adds should not be understated or ignored. As the number of virtualized servers and the mappings to their networked storage volumes grow, successfully mapping these storage volumes to each individual server manually can become both tedious and extremely difficult to complete successfully.

This also simplifies the occasional situations where virtualized servers are removed from the environment or replaced. Normally these tasks add entirely heightened levels of complexity to storage management but which this Autonomic Groups feature handles like any of the other tasks and accomplishes with the issuance of a single command.

Server virtualization is creating highly dynamic environments that can also become equally complex if organizations do not make tactical moves to mitigate this growing complexity. Today's release of the Autonomic Groups feature for 3PAR InServ Storage Servers strikes a blow to server virtualization complexity. It makes the initial and ongoing managing and provisioning of storage in virtualized environments easier and faster to accomplish while at the same time removing the risk and complexity that can quickly become pervasive in these environments.

A 2500x Efficiency Factor Highlights Differences Between Competing Thin Provisioning Solutions

2009-11-04T11:00:00Z

This past summer a rather humorous video debate broke out on the StorageRap and RupturedMonkey.com blog sites. The debate centered on the 42 MB page size of Hitachi Data Systems (HDS) Dynamic Provisioning feature (HDS's implementation of thin provisioning) and the 16K block size of the 3PAR InServ Storage Server thin provisioning technology.

Marc Farley claimed in a StorageRap blog post that HDS's thin provisioning technology is terribly inefficient in its design. Rupturedmonkey.com's Nigel Poulton fired back by saying that Farley's claims did not paint a complete picture. While both Farley's and Poulton's videos generated a chuckle from me, the thin provisioning technology that they talk about and upon which users are betting their business is no laughing matter.

To briefly recap the exchange between Farley and Poulton, Farley posted a blog on his StorageRap website on June 25, 2009, that claimed the 42 MB page size of HDS's Dynamic Provisioning was hugely inefficient when compared to the 16K page sizes of 3PAR's Thin Built In™ design. Within hours after Farley's post, Poulton posted a comment on Farley's blog and then filmed and posted a two-part video blog on his own site that explained why HDS's 42 MB page size was really a better solution for enterprises than 3PAR's 16K page size.

Both of these individuals had good arguments as to why one thin provisioning solution was better than the other. Farley's illustration in his video of how HDS carved out a 42 MB page file for new writes from each application server struck a nerve with me. For instance, if you have four application servers and they each send a net new write to the HDS storage system, HDS using Dynamic Provisioning will consume over 160 MB of storage while the 3PAR InServ Storage Server would only consume 64K of storage (i.e. - 3PAR's approach is 2,500x more efficient than HDS's approach on the initial write).

However I had to respect Poulton's counterpoint in his blog that subsequent writes from any of those respective application servers would not consume any additional blocks of storage on the HDS array. Rather they would continue to write to that initial 42 MB block of allocated storage on the HDS storage system and would not consume another 42 MB of storage until that first 42 MB of storage is filled up.

In the case of 3PAR, it would continue to allocate new 16K blocks as additional writes from the same application server arrive. So while 3PAR is arguably more efficient in how it initially allocates storage, the major differences between how the 3PAR and HDS storage systems manage storage capacity appear to primarily surface during the initial write and then when it fills up. Theoretically it is only during those two times that HDS's Dynamic Provisioning becomes grossly inefficient.

However an assumption being made by Poulton in the case of HDS is that subsequent writes from the same application server to the HDS thinly provisioned storage volume are always made to the original or subsequent 42 MB block of storage. This is not a safe assumption.

One of the wildcards in implementing thin provisioning is the type of operating system and its associated file system that are used in conjunction with the thinly provisioned volume. Some file systems when they are created write metadata in 8 MB, 10 MB, 52 MB and other larger intervals up to about 128 MB on the thinly provisioned volume. File systems that do this include AIX JFS, HP-UX HFS, Linux XFS and Solaris UFS. (This is illustrated on a table that appears on page 30 of the August 2009 HP StorageWorks XP24000/XP20000 Thin Provisioning Software User's Guide. The HP StorageWorks XP24000/XP20000 is HP's version of the HDS USP.)

So how does this file system behavior impact 3PAR and HDS/HP? Like anything else, it depends. If users are connecting Windows 2003 or Windows 2008 servers to either of these storage systems, 3PAR will still thinly provision storage more efficiently than HDS/HP. While initial utilization differences may be small, they will pile up in 3PAR's favor as application server IO's initiate 16K capacity increments versus HDS' 42MB. In addition, recent thin-related announcements from 3PAR suggest it aims to offer a significant advantage going forward.

However enterprise organizations increasingly measure their storage utilization in the tens and even hundreds of TBs, especially among the larger enterprise environments where AIX, Solaris, Linux and HP-UX systems reside and for which 3PAR, HDS and HP are many times gunning. It is in these environments that the efficiency of 3PAR's thin provisioning technology indisputably overshadows HDS and HP when file systems are initialized.

3PAR can do nothing to stop the file system from writing metadata at specific intervals but it can and does mitigate the impact that these file systems have on a thinly provisioned volumes by only consuming 16K of physical capacity as these file system metadata writes occur. In the case of the HDS and HP systems, they can do nothing. They will consume another 42 MB of capacity and, as the table in the HP User Guide points out, the entire size of the thinly provisioned volume is in some cases consumed.

Right now storage providers of thin provisioning are having a good time taking pot shots at one another in regards to their implementations of thin provisioning but for users, thin provisioning is serious business as choosing the right (or wrong) implementation of thin provisioning directly impacts an organization's bottom line.

In this case of 3PAR versus HDS/HP, it becomes apparent that 3PAR's more efficient implementation of thin provisioning will use less storage in every circumstance. However users measure success in dollars and cents and not saved (or unsaved) storage capacity. In this respect, 3PAR's thin provisioning technology has the edge as data grows in Windows environments over HDS and HP. But when it comes to AIX, HP-UX, Linux and Solaris environments, 3PAR has a financial edge that is measurable and definitive from the very start.

Cloud Storage Horror Stories Got You Spooked? Get Your Data Center Cloud Ready Now

2009-10-27T10:00:00Z

Cloud computing helps companies hide, intelligently allocate and take control of their IT infrastructure while also supplying users with the appropriate levels of application availability and performance. But, with this flexibility comes the inherent risks. One of the most pronounced risks is protecting the application data stored in the cloud because if you don't, cloud storage horror stories are sure to follow

This was brought clearly into focus a couple of weeks ago with the drama surrounding the status of the Sidekick data hosted by Microsoft. While the Sidekick data once thought to be permanently lost now appears to be on its way to recovery, the up to one million affected Sidekick users are not enjoying this drama as it plays out at all. If anything, it illustrates why data protection in the cloud should be a top priority for enterprises and service providers as outages create angry customers and result in lost revenue, current and future.

Yet protecting and recovering data in the cloud is problematic as the storage on which the data resides can be spread across multiple locations. This makes it difficult to keep the data in synch in the first place as synchronous mirroring (writing the data to both local and remote sites at the same time) is really not an option either as mirroring data over long distances can negatively impact application performance.

What are available are multi-site, multi-node options that offer zero I/O loss and enable fast recovery over long distances. But there is a catch. These multi-site, multi-node options are only available from high-end, monolithic storage arrays that come with high price tags, require substantial amounts of professional services to deploy and lengthy deployment windows. This often makes it impossible for cloud providers to consider these solutions when the costs and time are factored in.

In response to these new requirements, 3PAR today announced its new "Ready" technology that is geared towards helping data centers further bulletproof the data they store in the cloud. By using 3PAR InServ Storage Servers and its "Ready" technology as the backbone for their storage cloud, it can handle the workload and capacity associated with any applications that an organization chooses to host in the cloud as well as keep it fully protected and provide a disaster recovery strategy. Examples of the new capabilities that 3PAR's "Ready" technology offers includes:

Extends 3PAR's InServ Storage Server's Remote Copy Capabilities. In addition to existing synchronous mode and asynchronous periodic mode in its Remote Copy software, 3PAR now offers a third remote copy mode that links these two modes and provides long distance remote copy functionality without any I/O loss in a disaster scenario. Using this mode, a volume can be mirrored synchronously and asynchronously at the same time.

Using 3PAR's existing synch and asynch replication modes, Primary site A may synchronously copy to site B and asynchronously copy to site C. With this new mode, site B will keep track of the asynchronous snapshots occurring at site C. If there is a failure at the primary site A, site B will then push all the I/O since the last snapshot that occurred at site C. This allows site C to quickly catch up to the last acknowledged I/O of the primary site.

More impressively, this can working over distances of up to 3,000 miles without adding special extenders or converters to facilitate the long distance replication, deployment has been simplified to the point that professional services are an unnecessary expense, and organizations can mix and match 3PAR's enterprise and midrange arrays. For instance, organizations could deploy an InServ T800 at the primary site and then a smaller InServ F400 at the secondary site. This provides businesses a multi-site DR at half the cost of other solutions in a fraction of the time.

Eliminates Write-through Mode during Controller Failures. A controller or cache board failure on a storage array will typically result in a loss of half of the computing resources as well as a loss of the write cache. On the application side, this can have up to a 70% performance impact on the storage system and drastically affect application service levels.

To keep this from occurring on 3PAR's, 3PAR's new persistent cache will re-mirror itself across the surviving nodes. The write cache (the dirty cache that has not been flushed to disk) then gets re-mirrored so the application servers can continue to drive I/O to the array without the array going into write-through mode or compromising service levels.

Persistent Cache, in combination with the Mesh-Active controller architecture, means greater assurance that application and virtual machine service levels can be cost-effectively maintained, no matter what the circumstances. While the configuration does require four controllers, all of 3PAR's T-Class arrays as well as its midrange F-Class F400 array support four controllers (or more).

RAID MP. 3PAR's new RAID MP (Multi-Parity) is based on Fast RAID 6 and Rapid RAID Rebuild technology within its Gen 2 ASIC. Like traditional RAID 6, RAID MP protects against double disk failure but the cool thing about RAID MP is that its overhead is the same as RAID 5 3+1 in the 6+2 mode and it also gives users increased performance (it comes within about 15% of RAID 10), while removing some of the rebuild time concerns that come with large capacity SATA disk drives.

Once again 3PAR is extending the capabilities of its InServ Storage Servers into areas that are top of mind among data center architects moving into cloud computing with solution objectives aimed at avoiding Sidekick-like scenarios. These latest advances in 3PAR's InServ Storage Servers effectively deliver more robust service level protection and DR for cloud storage providers that were unattainable and unaffordable for many enterprises and service providers. By making data protection and disaster recovery an integral part of their cloud storage solution as 3PAR now provides, enterprises and cloud storage providers are better able to address design objectives aimed at avoiding the potential for Sidekick horror stories of their own.

Zero is Thin Provisioning's New Calling Card

2009-10-12T12:00:00Z

No one likes to be thought of as a "zero" but that is exactly how 3PAR wants the industry to perceive the newest management features on its InServ storage systems. The adoption and integration of thin provisioning into 3PAR's storage systems already give it a competitive edge. But today's announcement of its new Thin Engine, a "zero" out algorithm that facilitates the adoption and preservation of thinly provisioned volumes in new and existing deployments may prompt even more companies to zero in on the savings that 3PAR's systems can deliver.

EMC, IBM, HP, Hitachi, and Dell are often touted as the primary enterprise storage vendors with up to 70% or more of the storage market depending on whose numbers you believe while others such as 3PAR make up about 2% of the total storage market. But that tide could be changing as noted in a recent article which states, "a comparison with EMC and other enterprise storage array competition implies that 3PAR is gaining market share."

So what's the reason for this? The article goes on to state, "The company's (3PAR's) virtualized InServ arrays compete for enterprise storage business with EMC Symmetrix, HDS USP, HP EVA and XP, and IBM's DS8000 and XIV products. InServ products stand out for their early and strong focus on thin-provisioning and their popularity with service providers."

It seems that more companies in this tough economic environment are realizing the benefits of using thin provisioning and the ability for a storage array to provision capacity on a pay-as-needed basis. It improves capacity utilization, simplifies storage administration costs and lowers cost of ownerships while cutting the huge up front costs associated with traditional monolithic storage systems.

The rub for many companies is that if they are just starting out with thin provisioning, that is not so difficult. Data centers can simply introduce 3PAR into their environments for new applications, allocate some thinly provisioned volumes to these applications and away they go only consuming capacity as new writes are done on the volumes. The more data centers can just slap in some thinly provisioned volumes and off they go--consuming capacity as new writes are done on the volumes. The difficult part is getting off of the fat volumes that they are already using and migrating to thin volumes so they can take advantage of thin provisioning, which can often be the majority of their deployed capacity. The challenge has been that a storage technology refresh requires a terabyte-for-terabyte replacement of capacity.

Since many companies have not had this luxury of starting off thin, 3PAR has, in today's announcement, released four new products that will make it easier for companies to adopt thin provisioning even if they are already using fat volumes and then stay thin. The significance of these new products means that a technology refresh of a 200TB installation might only require 50TB of new capacity. That's an impressive savings in an area, tech refresh, that has typically been left out of capacity reduction efforts. Also as part of the announcement, 3PAR is providing new industry direction for companies to help them build thin ecosystems within their environments.

Getting Thin. The goal of getting thin (not to be confused with simply starting out thin) requires converting fat volumes to thin volumes. Moving existing volumes to 3PAR's thin volumes requires block-level migrations but until now the problem with this approach is that every block on the source volume is copied whether or not it contains data. A tech refresh was precisely a terabyte-for-terabyte project.

This then leaves new volumes on 3PAR systems over provisioned even after the migration is complete.The announcement of today's 3PAR Thin Conversion product changes this. 3PAR leverages industry standard block migration technology but now performs an inline zero detection using its Gen2 ASIC and virtualizes unused space from the fat volume with its Thin Engine as volumes are copied to a 3PAR array. This analysis, done at wire speeds, maps free space and only lays down written data on disk. The end result is a tech refresh that occupies far less capacity, costs a lot less, uses less power, and takes up less space.

The fat to thin migration using 3PAR's Thin Conversion is a simple 2-step process. A quick check of the file systems and related volumes give you a quick sense of the capacity savings. Zeros are then written within the file system to any unused space which lets the 3PAR ASIC detect what blocks are used and unused. Then as data is ingested by the 3PAR InServ system, it will only map and lay down blocks that contain actual data. 3PAR has found this method works as fast as the migration occurs, usually just minutes for most volumes.

Staying Thin. Keeping volumes thin over time can be as difficult as getting them thin in the first place. What can occur in environments with thin unfriendly file systems is that applications write out new data and then quickly delete it. The writes consume the available capacity but when data is deleted, the file system may not efficiently reuse the freed blocks and are "stuck" in the thin volume. This is also seen in applications that have lots of transient data that is written and then quickly deleted, or in the deletion of very old snapshots.

To address this, 3PAR's new Thin Persistence and Thin Copy Reclamation features use the Thin Engine to reclaim these freed blocks when data is deleted. Thin Persistence is ideal for keeping your VMware VMs thin by running utilities like Windows SDelete or the equivalent inside your VMs on a regular or utilization-trigged basis. It can also keep your VMware datastores thin by using VMfktools to zero out the VMDKs of VMs you are about to delete. The Thin Engine in Thin Persistence reclaims the capacity automatically as these tools are run. It can also be used periodically like Thin Conversion where zeros are written across deleted space in the file system which in turn, frees the associated capacity.

Thin Persistence can be turned on and off on a volume by volume and can be scheduled to run whenever - weekly, monthly, quarterly, or immediately after other scheduled tasks take place which in turn triggers the ASIC in the 3PAR system to de-allocate the underlying capacity.

Thin Copy Reclamation is similar to Thin Persistence but is used on copy volumes. Because 3PAR maintains the copy space mapping, as data is deleted, the Thin Engine immediately knows about the deletion and can actually un-map the capacity as it occurs. This means, when a snapshot that's been around too long for example and has grown in size, the capacity can be immediately freed from thin copies and used elsewhere.

Demand Thin. About a year ago, 3PAR announced a partnership with Symantec to jointly develop a thin API to leverage standard SCSI commands that would allow a host file system to inform 3PAR at a volume level which blocks were freed when files were deleted. This approach, which relies on a standard SCSI command called WRITE-SAME, is mirrored in the T10 SCSI standard for thin provisioning.

To encourage the "thin data center," 3PAR is actively prodding its file system partners to take advantage of this new thin provisioning API. It should be no surprise that the very first product built on this emerging standard is Thin Reclamation for Veritas Storage Foundation, given 3PAR's work with Symantec on the API. This product will leverage the API for communication between the VxFS file system and the Thin Engine within the 3PAR array so that as files are deleted on 3PAR thin volumes, Storage Foundation will inventory the block list and then push that block list to the 3PAR array which it will then use to free up that capacity.

In this economic environment everyone is looking for new ways to add value without sacrificing the features they need. Today's Thin Conversion, Thin Copy Reclamation, Thin Persistence, and Thin Reclamation announcements from 3PAR do so in a most innovative and simple way. By providing a utility that zeros out blocks that no longer contain data, not only does it contribute to saving money and easily recovering unused storage on thinly provisioned volumes, it gives companies a whole new way to look at the value of zero.

Download this blog as a DCIG Brief.

One Service Provider Says Goodbye to FC Disk Drives And Hello to ATA for its Oracle Production Workloads

2009-09-23T10:00:00Z

Storage system providers sometimes allege that ATA disk drive performance and reliability is no match for FC disk drives and ATA disk drives should certainly never be used in conjunction with enterprise applications such as Oracle databases. But when the cost of SATA come in at about half the price as FC disk drives, it behooves organizations to take a hard look to see if they can justify making the switch if it can be done without compromising the availability, performance and reliability of the data that it hosts.

ATA drives are breaking out of their initial deployments in disk-to-disk backups environments where FC disk drives are too costly. Through the implementation of FC-to-SATA bridges, ATA has emerged as a viable storage architecture alternative that can now support production applications. Of course, this may make sense when hosting file server data but is it still such a good idea when hosting production Oracle database applications on this tier of disk?
In these situations, one has to balance the risk versus reward of how well these ATA disk drives will perform under these types of application loads. To get an answer, I turn to a recent interview that DCIG conducted with Anthony Tarone, Director of Service Operations for CEDAR Document Technologies, and what his experiences were when he made such a switch..

CEDAR Document Technologies requires a lot of I/O performance from its backend storage system. Daily it receives anywhere from a few MBs of data from its smaller customers to hundreds of GBs of data from its larger customers. Once it receives the data, it validates, extracts, transforms, loads, and generates another 10-20 GBs of data to produce and electronically deliver the various documents for its customers that they then archive.

Further, CEDAR supports a growing base of customers using secure single sign on (SSO) transactions, has about 140 million documents in Oracle, and often retains archives for customers for 6-8 years. In a nutshell, CEDAR had a lot of different transaction types and storage options to manage. This environment makes it an ideal candidate to evaluate the performance and reliability of ATA-based disk drives in a mission critical environment.

The need for CEDAR to make a decision on ATA came to a head in 2007. It had been running its small customer base on 8TBs of direct attached storage plus some NFS for file sharing. However customer growth was pushing CEDAR beyond what its current infrastructure could support. After evaluating a variety of technologies and vendors, CEDAR made the investment in a 3PAR InServ storage system that supported a number of technologies, including ATA or "Nearline" drives.

The InServ storage systems gave CEDAR greater processing capacity and easier scalability on the storage layer as well as new freedom to utilize 3PAR's virtualization technology that distributed chunklets of data across its disks. Because of rapid growth and an unclear picture of future I/O loads with new customers and applications coming online, CEDAR opted for Nearline drives instead of the more expensive FC drives.

This may seem like a risky proposition to some but the decision was actually grounded in sound reasoning. The distributed nature of the 3PAR architecture provides wide-striping and improved I/O over a standard ATA-based array, log file write-performance that approaches the millisecond-range and the ability to plug in faster FC disks later as they needed proved to be a good decision for CEDAR. 3PAR calls such a solution "Nearline for Online," that is the ability to run demanding primary storage applications on a cost-optimized (ATA-class) storage tier. Contributing to its decision to select 3PAR, other vendors were seeking firm performance and throughput estimates for array sizing that was difficult for CEDAR to provide and was unnecessary for the 3PAR configuration which touts an ability to handle unplanned growth and changing workloads. Also, another system would have locked CEDAR into an architecture that was not as flexible, distributed, or capable of handling a mix of ATA and FC drives without taking a performance hit.

As CEDAR's customer base continues to grow, and it adds more disks to accommodate storage requirements, performance continues to get better as I/O is distributed across more disks. The point is that CEDAR was not been able to justify nor did it have a real need to pay for traditional FC disks which are 2 - 3X more expensive than the nearline drives. Instead it was able to implement a low cost hardware solution and still support Oracle which has enabled CEDAR to successfully continue to store over 140 million documents and service its customers in a timely manner.

At the end of the day, the combination of flexibility (and budget) pushed CEDAR towards 3PAR. Being a service provider, CEDAR is continually concerned with building in high availability for the services it provides at the best possible cost. With 3PAR, CEDAR has not had to purchase expensive FC disk, has not had a single outage since deployment 2 ½ years ago and, because of 3PARs technologies and virtualization platform, CEDAR has been able to use the ATA-class drives for all production services without negatively impacting performance, SLAs, or experiencing any risk of any sort; even to the extent of running high-performance Oracle databases.

Using 3PAR storage and its service and support, CEDAR is well positioned to continue adding new customers without fear of a breakdown in its storage infrastructure. The question of FC or ATA disk is now a moot point as 3PAR makes use of, and has for a while, a FC-to-SATA bridge that enables high-capacity ATA drives to be integrated into its storage arrays that provides customers like CEDAR a low-cost but highly available storage infrastructure for even the most demanding application loads.

A Half Million Dollars in Savings - Good; A Fully Supported Storage Solution - Priceless

2009-09-16T10:00:00Z

At a time where vendors are positioning savings "guarantees" to draw attention to their storage offering, it is refreshing to see a storage user actually tout substantive savings just by switching to 3PAR. This was accomplished recently done by CEDAR Document Technologies who announced it saved a half a million dollars, improved performance, experienced a 5x increase in transaction volumes and avoided $250,000 in administrative costs just by switching to 3PAR's InServ storage systems. But to find out just how things were going with CEDAR now that it had used 3PAR for awhile, DCIG checked it with its Director of Service Operations to see what other benefits it was deriving from its 3PAR storage systems.

CEDAR Document Technologies is a service provider of print and electronic customer communications--helping customers communicate better with their customers. In general, CEDAR takes customer data that would normally be taken to a print or mail shop and instead decomposes it into XML, reformats the data, enhances the XML with rules or additional elements, and then manages different delivery channels including postal mail, secure push, email, portal, text, etc. as well as manage the archival of these documents.

As CEDAR's customer base began growing, not only in quantity of customers but in the amount of data being received on a nightly base, their current direct attached storage was unable to handle both the amount of data and the processing required. Small and legacy customers were not the issue; it was the new customers that began sending anywhere from 50-100GB of data.

CEDAR needed a new storage infrastructure that would continue to grow as the company took on these larger customers. 3PAR was heavily looked at and ultimately chosen because they were the only vendor that could through its virtualization layer supply storage that would continue to expand and provide the performance needed to meet CEDAR's rapidly growing customer base as well as let them remain agile as their operating environment changed. Here's what DCIG learned after talking with Anthony Tarone, Director of Service Operations for CEDAR document technologies.

DCIG: Can you first give us an idea of what your customer base looks like in relation to the amount of storage you purchased from 3PAR?

Tony: As a service provider CEDAR deals with lots of data. Our smaller customers may only send us a few MB a day which is somewhat unnoticeable in the grand scheme of things. However a mid-size customer may send us 10-20 GB of data per night which, after going through a multi-step workflow process, explodes out to 60-70 GB. Then our biggest customers may send us 50-100 GB of data per cycle which can see similar explosions in data growth.

Prior going to 3PAR, CEDAR used to process this data with direct attached storage and use NFS for file sharing. It was good for our smaller customers but we needed a new infrastructure for the new breed of customer we were acquiring to handle the greater processing capacity and scalability on the storage layer that their data loads created. Back then we only had 8TB of direct attached storage. Now, after rebuilding our entire infrastructure in a collocation site, we have 54TB in production and 24TB in development, all of which is 3PAR.

DCIG: How has the move to 3PAR impacted your total cost for a storage solution? We've read in the press release about the half million dollar reduction in total storage costs but how did you achieve this?

Tony: From a purchasing standpoint, the initial quotes we received from other vendors were well above and beyond what it cost for the 3PAR solution. Using 3PAR, we not only were able to get storage requirements but our corporate development, and QA environments as well.

Additionally we have not spent a penny on professional services with 3PAR, and it is not because we have not done anything. 3PAR's support team is absolutely outstanding, especially their higher level support at corporate as well as their sales engineer here in Atlanta. If we ever have any questions, are trying something out, or have a problem we can, depending on his availability, count on our local SE to pop over, have some lunch with us, check our configuration, maybe consult with top level support teams, and get us an answer.

3PAR does this all from a good customer service standpoint and they're not hitting us with a quote for professional services every 15 minutes. Further, as a service provider our main goal is to first build in availability throughout our infrastructure; with 3PAR we have not had a single issue or outage since deployment 2 ½ years ago. This is priceless.

DCIG: It sounds like a lot of your cost savings came from the upfront purchase of the unit as well as decreased support costs. What about performance and the need to manage all that storage?

Tony: That's exactly right. I'm not getting hit with professional service fees. As far as management goes, we've been able to manage the storage with our current staff. I haven't had to hire a SAN administrator or any dedicated staff because 3PARs tools are easy enough to manage that we've been able to do everything with very little training from our SE.

The performance has been exceptional and even improving as we add more customers because of the way 3PAR spreads chunklets out across disks to the point where I still haven't been able to justify or have a real need to pay for fiber channel disk which are vastly more expensive then the ones we have now. Because of 3PAR's technologies and virtualization, I've been able to use their Nearline fiber channel ATA drives (FATA) and serial ATA drives (SATA) for all of my production services without impacting performance, SLAs, or creating any real risk of any sort.

DCIG: That brings up an interesting point. You took a risk where many data centers would have opted for the higher priced regular FC drives. Have you noticed any increase in costs or service problems associated with the nearline drives you selected? Aren't they supposed to fail more often?

Tony: The drives have actually been quite nice. 3PAR support can proactively detect when a drive is failing and coordinates everything. 3PAR moves data off of the offending magazine, schedules someone (usually the next day) to come out and replace the magazine, and then redistributes the data back over it. The drives haven't been an issue at all. Best of all there isn't a fee for the disk swap. That is all part of the annual support, and it's my understanding that as long as we keep it under support 3PAR will keep changing disks out. The whole process has been very seamless, painless, and literally a non-event for me.

DCIG: Have you done any upgrades to your existing system and have the upgrades gone as anticipated? Also, have you performed any dynamic upgrades?

Tony: We've done some upgrades although I'm more old-school and usually only do those within my dedicated maintenance window; I just try to be safe. In fact we have an upgrade coming up where we will be cross connecting all of our redundant power which used to be on a single power plant.

3PAR is a part of that because we have a 30AMP hardwired into our 3PAR unit. We'll be working with 3PAR on the maintenance window. Our local 3PAR SE (support engineer) will coordinate 3PAR activities so that when we unplug it they'll be checking it out, getting some alerts from our system, and managing the process. 3PAR has always been completely accommodating to our schedule by either having their people onsite or on the phone with us. They have never forced us to fit into their schedule.

DCIG Final Thoughts: Purchasing a storage solution is more complicated than just answering questions about capacity and up-front costs. Organizations need to take a hard look at what their problems are, investigate competing solutions, identify the benefits and understand other hidden storage system costs, such as maintenance, support and the timeliness of responses to customer questions. After talking with CEDAR, it became quite clear that 3PAR saves money and delivers a better storage solution which is always good. However 3PAR's continuing willingness to step up to the plate and provide exceptional ongoing customer service and support makes CEDAR'S decision to select 3PAR over competing solutions appear priceless.

Are We Really Ready for Deduplication of Databases on Primary Storage? I Still Say No

2009-08-13T10:52:12Z

In a recent blog post entitled Deduplication of Databases on Primary Storage Just Rubs Me the Wrong Way I received some great comments, questions and even a ding. Because of the nature and depth of the comments and questions, however, I felt it only appropriate to produce a follow up to that post and help explain a few things brought up in those comments. In particular I wanted to address questions raised by Matt Simmons and Mike Dutch.

Matt Simmons said: Please correct me if I'm wrong, but isn't there a performance hit for doing random access on deduplicated data? I don't see how that couldn't be the case, since every data request would have to be looked up in a table. Even if the entire data store is stored in cache, that's still a lot of latency just from lookups. Of course, I could be misinformed as to how dedupe works.

Koopmann says: Intuitively, we all know that adding additional code, instructions, or hardware will add processing overhead. What we all want to know is, "How much?" More often then not we ask ourselves or look for a bigger box that will hide the performance hit

Since deduplication on primary storage is semi-new, performance numbers and real-life scenarios are hard to come by and even a number of deduplication vendors are forthright in saying their deduplication is intended for backup environments. However I think we can safely say though that deduplication on primary storage is much more difficult than deduplication of backup data. So I'd like to just leave a few quotes on what some of the experts in the field say:

In regards to deduplication of primary storage for backups I'd like to direct you to a quote I found from Larry Freeman, Senior Marketing Manager Storage Efficiency Solutions at NetApp and Dr Dedup himself. On June 18, 2009, Dr Dedup says: "If your application or system is extremely performance-sensitive, don't run dedupe". Now I personally wouldn't take such a hard and fast stance here and I'd encourage you to read the rest of the thread as there is some good information on how NetApp's deduplication works along with performance data.
In a Storage Switzerland, LLC LAB REPORT: Deduplication of Primary Storage, Senior Analyst George Crump states that "As I have stated in other articles, deduplication by itself has limited value on primary storage" and "works against active online and near line data (like databases), where the occurrence of duplicate data is unlikely. This is in contrast to backup data, where the same full backup runs every week and the chances of duplication are fairly high".
In a SearchStorage.com article, Users turn data reduction focus to primary storage, Senior News Director Dave Raffo asked the questions 'Is data deduplication a good fit for primary storage?' Data Domain CEO Frank Slootman said not only do "We try to use the term 'primary storage' carefully," but "If you have data that is really hot, has an extremely high change rate, like transactional data, there's no sense deduplicating it."

The problem here isn't really about how much of a performance impact you will see but more along the lines of, "Can you really measure the performance impact?" and "Does anyone really care or perceive it?" I've often found in the database shops that I have worked in that there are very few people that know how to relate storage performance to database performance. Just ask your DBA if they know how many IOPS they are getting and if they require additional spindles to help improve SQL performance. (More than likely they will roll their eyes.)

Mike Dutch said: Isn't database normalization/record linkage essentially concerned with "information deduplication" focused on ensuring "good" query results rather than "data deduplication" which is focused on improving storage and bandwidth consumption? I tend to think of "data deduplication" as being part of the data storage process as it is similar in many ways to "just another file system"... "go here to get the next set of bytes".

Since most data deduplication procedures store data compressed, it must be uncompressed and that network bandwidth/latency issues may result in undesirable performance impacts for some data. However, this doesn't mean that you shouldn't use data dedupe on some databases (for example, copies of a database used for test purposes) or on other types of primary storage data that are not overly performance-sensitive.

Koopmann says: Yes, database normalization is concerned with developing structures that will not only help ensure data quality but also help ensure that an application is able to manipulate data easily. While a byproduct of normalization is the reduction of redundant data let's stop right there and also realize that normalization does not eliminate all redundant data.

Again, in a purist sense, data modeling, when done effectively (or compulsively depending on your perspective), will push down all redundant data to lookup tables--leaving UIDs as pointers to that lookup data. At this point duplicate data is completely eliminated except for the system generated UIDs that you would be hard pressed to note any storage gains through deduplication.

Granted, most databases are not modeled this way so looking at how your particular database is architected and works is key to breaking the deduplication mystery (with vendor help). For Oracle, databases structures (table and index designs) will share the same blocks but the actual data/records/rows that are stored in the data blocks are likely to be different.

But it does depend on the data model and transaction patterns (INSERTs, UPDATEs, DELETEs). This means having deduplication at the file level surely won't work and deduplication on individual blocks probably won't see duplicates. This leaves us with deduplication at the byte level within and across all data blocks in all SEGMENTs of data files. Depending on your dedup vendor this solution may be beneficial but you, will certainly need to verify the solution.

Surely you can use deduplication on some databases as Mike suggests and hit it square on the nose. The key term used here by Mike is "copies of". Many databases have copies of data that may be for backup, test, historical, or offloading query functions. If your database has many of these "copies of" data, you could look at the benefits of deduplication, or perhaps consider non-duplicative "thin" snapshots. Non-duplicative snapshots refer to the fact that changed data is never duplicated across a group of snapshots versus copying changed data across all snapshot copies. This particular technology is well-proven in primary storage applications.

Again, I have nothing against deduplication when used appropriately and other factors are uncontrollable. But in a properly architected database I am still skeptical about the fit. Databases are just too dynamic with temporary sort, rollback, and redo areas and high transaction rates that make me question what there could possibly be to dedup in the first place.

Secondly, many database practitioners are too far removed from the performance implications that revolve around storage already that they would certainly not be able to handle an additional layer of abstraction. So for now, I hold steady on my statement that a more viable alternative to current reduction and stabilization of storage acquisitions within database environments would be to deploy thin provisioning.

Using proven and industry leading thin provisioning and non-duplicative thin copy technologies from a vendor such as 3PAR allows databases to allocate just-enough and just-in-time storage--relieving IT from both having to watch and then add or remove storage. Thin provisioning and thin copy are data reduction technologies developed for primary storage applications, and therefore addresses the desire for capacity efficiency without the performance impact and deployment mystery associated with today's storage deduplication technology.

Deduplication of Databases on Primary Storage Just Rubs Me the Wrong Way

2009-07-27T11:12:53Z

First, let me set the record straight. I have nothing against deduplication. Deduplication is a proven technology that has many benefits within an IT data center--providing improved data protection and reduced costs associated with storage acquisition.

But let's remember, deduplication is a technology that basically reduces or eliminates redundant data--leaving a unique data set. And if you've been keeping up to date on your Net reading, deduplication, which in the past has been primarily used for archive storage, is starting to be heard when talking about primary storage. Now I can't talk for all the 100's of different applications running out there but deduplication for applications utilizing primary storage with a database backend just rubs me the wrong way.

From a DBA purist perspective, database design, aka Database Normalization, is undertaken to help eliminate redundancies and facilitate the quick retrieval of information. Place this against the back-drop of deduplication and you might soon begin to wonder about the effectiveness of your database design.

Sure there are instances where duplicate data in a database exists and its reduction would be beneficial; instances such as tables that keep track of male/female, city/state, or yes/no data. There are even features to help with redundant/duplicate data. For instance, Oracle has embraced CLUSTERs and DEDUPLICATE options for building tables as well as bitmap indexing to help manage duplicate data. But these options are not the norm and have very narrow purposes. The key point here is that databases, by definition, are meant to help eliminate redundancies, not produce or propagate them.

Personally, if someone told me that they could reduce my database storage footprint by 50% I'd begin to worry about data quality within my database. Reducing the storage requirements for databases, through deduplication, really just puts a Band-Aid on this problem and doesn't address the real issues.

Real data quality and elimination of duplicate data are achieved by starting with a pristine data model, quality data, and validating an application will manipulate the data properly. Data modelers, along with DBAs, are responsible for validating schema design and ensuring various indexes, relationships, and constraints are in place to help improve the quality of data--ultimately helping to eliminate duplicate data.

Maybe if we spent more time on data quality, IT, as reported by Gartner, wouldn't be facing the extreme failure of data warehouse projects. And while Ted Friedman, principal analyst at Gartner, clearly states that the reason for a lack of data quality is most companies focus only on the "identifying, extracting, and loading of data into the warehouse but do not take the time to assess quality/" I am certain this lack of quality of these extract, transform, and load (ETL) processes are jeopardized through poor design.

Until a new form of deduplication technology comes out, deduplication on primary storage for a database environments just doesn't mix. Deduplication is still a viable technology for archival and backup, but only if you understand your data. For instance, you might want to use deduplication for more static tables pages during backup but you wouldn't think of using deduplication on Oracle archive redo logs since redo logs contain the real-time block changes to data. Granted, this could change with new technology that would recognize those large hot data blocks that have small changes but we just aren't there yet.

A more viable alternative to current reduction and stabilization of storage acquisitions within database environments would be to deploy thin provisioning. Using thin provisioning from a vendor such as 3PAR, databases can be allocated with just-enough and just-in-time storage--relieving IT from both having to watch and then add or remove storage. Thin provisioning is the one data reduction technology developed for primary storage applications, and therefore addresses the desire for capacity efficiency without the performance impact associated with today's storage deduplication technology--which was really developed for archive storage applications.

Storage management has always been the bane of DBAs so deduplication will probably move up the storage stack and end up in primary storage some day but that day is not today. In the meantime, using software like 3PAR's thin provisioning, DBAs can now provision once and only pay for what they use when they use it which eliminates the tedious, manual, and error prone tasks associated with database storage management while still keeping their data stores under control.

Building a Cloud-Based Computing Infrastructure: Where Star Trek and a London Data Center Meet

2009-07-08T10:00:00Z

Let's face it - the concepts behind cloud computing and cloud storage are just plain cool. The whole idea that an organization can move any application to any hardware platform and pro-actively meet the constantly changing needs of an application is more akin to something you might see in an episode of Star Trek than something you might find in today's data centers. Yet cloud computing and cloud storage are becoming a part of today's reality for tech-savvy data centers because these feature are not just cool, they ease the complexity of common data center management tasks, drive down costs and increase profitability.

Recently I had the opportunity to speak to Jason Reid, Carrenza Hosting's Director of Information Services, who is implementing "Star Trek"-like cloud computing and cloud storage technologies into his data center. As he explains it, cloud computing and cloud storage are cool marketing terms and reflect what he hopes to accomplish someday but at this point, he is still putting a foundation in place to deliver that functionality.

He prefers to call the infrastructure that he is creating now a "utility". While "utility" may not sound as sexy as "cloud", he wants to take his organization's existing and new computing resources and use them as efficiently as possible. By creating an infrastructure where all of his resources - server, network and storage - are virtualized, he can raise the level of conversation that he has with his clients so they do not just discuss what type or size of server or storage they need to support a specific application.

He says, "I want my customers to tell me how many CPU cycles, how much RAM and how much disk they need and then I will deliver it without getting into the specifics of what type of hardware is going to support the application. This is creating an infrastructure that operates like a utility."

Carrenza found that as it virtualized its infrastructure and managed it more like a utility, it realized just how low its levels of utilization had been as a physical environment. This was costing Carrenza in a number of different ways but a big one was in its inefficient use of data center space.

Carrenza's data center is located in London, England, where the per square foot cost of data center space is very high. However by using this space more efficiently, it had a huge impact on Carrenza's profitability. This resulted in it being able to lower the price it charged to its existing clients as well as made it more competitive as it sought to attract new clients.

Because Carrenza operates as a managed hosting company, Reid regularly has conversations with clients - new and old. During these conversations, he encounters a wide range of sophistication among Carrenza's clients where some customers are extremely sophisticated while others are less aware of how Carrenza's infrastructure is architected.

But once he explains to them how Carrenza has re-architected its data center to act more like a utility, their interest in the technology wanes. Reid comments, "Even technically savvy clients seem to become less interested in what is going on underneath the covers once we give them the impression that we understand what it is they are trying to do."

Carrenza still does receive requests for it to configure its infrastructure in certain ways to meet specific application requirements. This can pose a problem in the eyes of some customers accustomed to working in a physical environment since Carrenza uses 3PAR InServ Storage Systems. Its virtualization technology efficiently provisions storage and the data that is placed on the system is striped across all or most of the disk drives in the system.

The 3PAR storage system functions as a shared storage pool, which traditional wisdom would suggest creates the potential for contention. 3PAR addresses the usual concerns of quality of service on a shared storage platform with a combination of the wide-striping mentioned above as well as mixed workload support--the ability to run both transactional and sequential applications without the usual impact seen on traditional arrays. The upside is that all the storage is shared, performance needs are met, and service becomes much more economical for Carrenza's clients.

To counter these types of concerns, Reid told me that even as he and I spoke he had a demo running for a prospective client that had requested dedicated disk on a 3PAR system for their application. However, the disk had not yet arrived from 3PAR so he went ahead and put the customer's application in the shared disk pool on the 3PAR storage system.

After letting it run for a period of time and taking some performance benchmarks, the prospective client found that the performance of the application on 3PAR's shared disk pool was "quite good" so the client is re-assessing what they really need right now. To get dedicated disk in Carrenza's environment means they have to pay extra whereas if they leave the application in the shared disk pool, their storage costs decrease.

Everyone loves the sound of "cloud based infrastructure" if no other reason than it sounds cool and hip and the possibilities of what one can do with a virtualized infrastructure are seemingly endless. But the reality is that companies like Carrenza that are putting in cloud based computing infrastructures are doing so to solve real-world problems with no aspirations for guest appearances in upcoming Star Trek episodes. If anything, most are just happy to use cloud computing to solve today's business and technical problems so they can simply get home in time to watch that evening's episode of Star Trek on TV.

In a previous blog, I spoke to Jason how he was creating a more flexible IT infrastructure so Carrenza could meet a client's forecasted spike in the need for computing and storage resources with minimal or no disruption to the client's application service levels.

Cloud Storage Architecture Gives MSP New Found Flexibility to Respond to Spikes in Application Requirements

2009-07-01T10:00:00Z

One time occurrences that are accompanied by spikes in capacity and performance requirements are the bane of data centers. While many organizations can excuse IT for their inability to respond to unexpected one-time or occasional demands, perceptions and attitudes change when organizations know a heavy load is coming and IT cannot adeptly respond. It is this type of challenge that Carrenza Hosting, a managed hosting company based in London, England, intended to solve when it began to start down the path of adopting cloud-based computing and storage services. Part of its intent was to create a more flexible infrastructure so it could meet a client's forecasted spike in the need for computing and storage resources with minimal or no disruption to the client's application service levels.

The impetus behind my conversation with Carrenza's Director of Information Services, Jason Reid, was sparked by a recent joint press release from 3PAR and Carrenza. 3PAR recently made a technology donation to Carrenza in support of Comic Relief's annual charity event, Red Nose Day. On the surface, 3PAR's contribution seemed well-meaning and innocuous enough but taking advantage of a contribution of storage hardware is not as always as easy as it sounds.

Having previously working in an enterprise data center, installing storage, migrating the application to and from it and then decommissioning the storage once an event is over to return it can be a laborious and time-consuming process - so much so that the effort involved can outweigh whatever benefits the technology was supposed to provide. So to better understand why Carrenza needed this donation and how (and if) it could took advantage of 3PAR's contribution without undue labor I spoke to Jason Reid, Carrenza's Director of Information Services.

Reid explained that Red Nose Day is one of the largest one day charity events in the UK that takes in about $80 million US dollars. Comic Relief, the organization that runs Red Nose Day, normally has a platform that is quite small as there are normally not that many transactions occurring on a day-to-day basis.

But when the Red Nose Day occurs, it goes from needing a very small platform to one that is very large. To support this once-a-year event, Carrenza works with its technology partners (3PAR, Cisco, and HP) to provide extra capacity for the Red Nose application platform just for the duration of the event. Reid says, "Comic Relief does not need this huge platform sitting there for the rest of the year consuming power, costing money and generally being completely unused."

To enable the movement of the application from a small, less-powerful hardware platform to one that can handle these extra transactions, the entire infrastructure behind the Red Nose application is virtualized and highly redundant. So to move it, Carrenza used VMware's vMotion feature to bring up another instance of the application on a more robust hardware platform, migrate the data and then shut down the old instance.

Carrenza's Reid leveraged 3PAR's storage contribution in a slightly different way as 3PAR's contribution came in two forms. 3PAR first contributed extra disk that went into one of Carrenza's 3PAR storage arrays at Carrenza's primary site. This storage was used for the day of the event as it was needed to support the Oracle database back end of the Red Nose application.

3PAR's other technology contribution was made in the form of a second array that was installed at Carrenza's secondary site. This was needed to support an Oracle RAC stretch cluster that was put in place to support the application on the day of the event. Since this cluster configuration has severe distance limitations and Carrenza did not have sufficient capacity at its secondary site to handle the requirements of the Red Nose application, it needed the secondary array from 3PAR to support these requirements.

But most impressive was Reid's testimony as to the ease of implementing the new 3PAR storage and then decommissioning it when the event was complete. Reid estimated that it took at most a day and a half for one of his engineers to configure the storage and get the Red Nose application up and running on the new server and storage hardware, including the syncing of the data between the primary and secondary sites. He says, "There was no impact on the night of the event, it ran perfectly. There is not much more to say than that. We did not have any problems nor did we encounter any performance bottlenecks. Then once the event was over, the storage was just as easy to decommission as it was to setup."

Storage management and data migrations are two of the biggest obstacles that companies now face as they look to respond to occasional or one-time spikes in demand. However as Carrenza's experience with the 3PAR InServ Storage Systems during the Red Nose Day event shows, these obstacles can now be overcome. By using 3PAR as its back end storage platform, Carrenza was able to dynamically and quickly scale out compute, capacity and performance for a client application and, once the event was over, just as easily scale it back to its previous levels with minimal or no disruption to its client and without requiring heroic efforts on the part of its IT staff to accomplish this task.

But what makes this especially interesting is how Carrenza is leveraging its virtualized infrastructure to grow its business and become more competitive. In a forthcoming blog, I'll get into how Carrenza is leveraging 3PAR to deliver this more robust, virtualized infrastructure to lower its costs, set the foundation for the foundation for its future and meet the current and future needs of its clients.

Knowing What Tweaks You Can Make to the Microsoft Exchange Spreadsheet Calculator

2009-06-24T10:00:00Z

Plug-n-play - that's part of the idea behind the Microsoft Exchange Storage Calculator spreadsheet which provides organizations the general guidelines that they need when planning and configuring the storage that will support an instance of Microsoft Exchange. However one should not assume that this spreadsheet takes into account every possible variable regarding storage systems - it most certainly does not and says as much within the spreadsheet. So before making any tweaks to the spreadsheet, one must know something about the capabilities of the storage system that one will use to host Exchange and what changes one can confidently make to the spreadsheet without botching an Exchange implementation.

The "Storage Design" page within the Exchange Calculator spreadsheet is the one that organizations will specifically want to take a hard look at when they start to plan for and configure the storage system that will host Exchange. At the top of this page, Microsoft provides some guidelines as to how much overhead a disk drive failure within a RAID group will introduce during rebuild times (35% for RAID 1/0 and 100% for RAID 5 or 6).

The motivation behind Microsoft providing these percentages is to assist organizations in appropriately sizing their storage system should a disk drive failure occur. These percentages ensure the storage system maintains optimal performance for Exchange even in the event of a disk drive failure. So as an organization interprets these percentages and applies it to their environment, this is the amount of extra capacity they must add to their storage system to maintain production levels of IOPs and throughput for Exchange.

In the case of RAID 1/0, an organization needs an additional one third (1/3) more capacity (35%) to ensure forecasted production levels of I/O are met during the rebuild of a failed disk drive. In cases where RAID 5 or RAID 6 are used on the storage system to support Exchange, organizations need to plan to implement twice the disk drives if they want to offset the 100% overhead that rebuilds of disk drives in these RAID configurations incur.

While these overhead percentages associated with these different RAID configurations appear logical on the surface (and they are), organizations should verify that these percentages are applicable to the storage system that they are looking to implement behind Exchange. Since Microsoft makes a number of assumptions in its spreadsheet about RAID technology, this is one area that organizations can look to tweak without negatively impacting their Exchange implementation.

Bill Plein, a storage architect with 3PAR's Strategic Accounts Engineering Team, explained that he works very closely with 3PAR customers to help them understand that they can satisfy all of Microsoft's best practices without needing to follow this particular detail on Microsoft's storage calculator spreadsheet. While 3PAR supports and follows all of Microsoft's best practices, he helps customers understand some recommendations around storage design are important for availability reasons and others for performance.

In the case of the 3PAR InServ Storage Server, it uses a wide striping technology so it does not have the same performance problems regarding the RAID rebuild overhead that Microsoft's recommendations are trying to address. Instead 3PAR spreads out its rebuilds across every drive in the system, not just to a single replacement drive.

So if a 3PAR InServ Storage Server has hundreds of drives in it and it looses one drive, almost every drive in the system get a little bit warmer because all of the drives start rebuilding the parity or the mirror chunklets that were lost when that one drive went down. Then when that drive is repaired, all of the data is leaked back to the new drive in a background process.

Since the drives participate one by one and restore the data in the background, 3PAR's rebuild time is fast and the performance penalty is extremely low - less than 5% if measurable at all. So even though Microsoft says allow for a 35% overhead on RAID 1/0 systems and 100% on RAID 5 and 6 systems, 3PAR can deliver the same reliability and performance using its wide striping technology as these other approaches with less than a 5% overhead on its system.

Plein concedes that 3PAR customers conservatively plan for a percentage higher than 5%. He says, "Even if the customer goes with an ultra-conservative ratio of 10 - 20% of overhead for RAID 5, 3PAR is basically selling its customers less disk as they do not need all of the Mirosoft-recommended spindles. Here is a place where 3PAR does it better and they can back off on the Microsoft recommendation."

In most organizations, Microsoft Exchange is too critical of an application both from a business and technical perspective to make any tweaks to Microsoft's recommended configurations without solid evidence that they will work as designed. However this is one example where Microsoft is over-compensating in its recommendation because it has to design its specifications to the lowest common denominator which, in this case, is traditional RAID arrays. Using 3PAR's wide striping technology in lieu of traditional arrays, the customer can be pretty confident that when the solution is rolled out there will be ample head room on the system without needing to purchase a lot of extra disk to accomplish that objective.

Satisfy Microsoft's Best Practices for Exchange Storage Configurations While Using Next Generation Storage Systems

2009-06-17T10:00:00Z

Best practices for configuring storage systems in enterprise Microsoft Exchange environments requires that they look beyond Microsoft's basic guidelines for how to configure their storage systems. While these guidelines may be sufficient when deploying Exchange in smaller shops where direct attached storage (DAS) or traditional array-based storage systems are the norm, enterprise organizations need to know when to look beyond these guidelines and tweak them when deploying Exchange on a next generation storage system like the 3PAR InServ Storage Server. The key to making these tweaks, however, is to satisfy Microsoft's best practices without negating the inherent benefits that storage systems like 3PAR provide.

As previously explained, one of 3PAR's key benefits is that it can disperse data throughout its entire storage system using its wide striping technology. In most cases, organizations will achieve even better performance metrics than what they saw when following Microsoft's recommendations for configuring storage. However because 3PAR Storage Servers use this wide striping technique (Marc Farley nicely illustrates how wide striping works on this blog post), organizations also should take advantage of 3PAR's template feature so they can concurrently use wide striping and still satisfy Microsoft's best practices for Exchange.

Bill Plein, a storage architect on 3PAR's Strategic Accounts Engineering Team, explained that there are different kinds of templates on a 3PAR storage system. One type of template is the volume template that can describe in great detail the layout of the proposed volume and even allows administrators to potentially lock down certain variables so they can't modify the template or change certain values. The value of these templates is that they only need to be created once and can then be re-used multiple times.

In the case of Exchange, organizations need to segregate the volumes on which they put their Exchange data and logs. So for example if an organization needs a 9:1 ratio between its Exchange data and logs, administrators can simply choose magazines or disks 0 - 8 from every tray or shelf of disks and assign those disks to the "Exchange Data" template and then choose magazine or disk 9 and put that disk into the "Exchange Log" template.

Configuring the 3PAR system this way enables organizations to then follow and implement best practices for Exchange storage configurations such as aligning volumes using the Windows 'diskpart' command or segregating data and log files on the storage system. Equally important, templates enables organizations to keep 3PAR's wide striping feature in effect since it stripes data across all of the disks in that template pool.

3PAR's autonomic provisioning feature also remains in play when templates are used. When Exchange needs more storage, administrators only have to instruct the 3PAR system to create a log or a data volume. When that request is submitted, 3PAR assigns the volume to the appropriate template and then the data is appropriately striped over these resources. Plein says, "At this point, organizations are automatically following Microsoft's best practices for Exchange since the policies are embedded into the templates that are part of the 3PAR storage system."

Just because enterprise organizations are placing their Exchange application on next generation storage systems does not mean that they have to sacrifice best practices, ease of administration or performance - they can still achieve all three of these objectives. Granted, to satisfy all of these requirements you can't just back up the truck, unload the 3PAR InServ Storage System, connect it to Exchange and expect it to be in compliance but neither is it an exceptionally arduous, ongoing task that it can become on other systems. A little upfront planning and work should enable enterprise organizations to satisfy Microsoft's best practices for Exchange storage configurations while still keeping their enterprise storage system supremely easy to configure and manage.

Microsoft's Best Practices for Exchange Storage Configurations often Assume DAS not SAN

2009-06-04T10:00:00Z

Microsoft provides a large number of guidelines for how to properly configure storage systems that are used in conjunction with Exchange implementations. But an area where Microsoft often still comes up short is in providing best practices for configuring Exchange in conjunction with today's next generation storage systems. Many of Microsoft's storage recommendations are based on the assumption that organizations are deploying either direct attached storage (DAS) or storage systems with traditional RAID architectures. But with next generation storage systems such as the 3PAR InServ Storage Server that deliver features such as wide striping, old rules for Exchange storage configurations do not always apply.

To better understand how to balance existing Microsoft recommendations for Exchange storage configurations with new capabilities on modern storage systems, I recently had the opportunity to discuss these issues with Bill Plein, a storage architect on 3PAR's Strategic Accounts Engineering Team. Bill serves as the liaison between 3PAR and its clients in their Exchange deployments and endeavors to provide storage configurations that best satisfy customer requirements for capacity and performance while still falling in line with Microsoft recommendations which is not always easy to do.

A number of the problems that Plein encounters in the field stem in part from assumptions originally made when Microsoft developed its best practices for storage configurations for its release of Exchange 2007. As part of that release, Microsoft tried to correct some complexity and performance issues that arose from Exchange 2003 deployments where there were numerous suboptimal disk configurations. In particular, smaller organizations were encountering issues when deploying Microsoft Exchange on SANs as they had few, if any, best practices for managing Exchange on SAN implementations in place.

To try to improve this situation, Microsoft came out with a very strong framework around DAS for those shops that had smaller storage configurations. The problem that Microsoft specifically was trying to solve was to dispel the notion that a SAN was necessary to support Exchange. Microsoft found that in these smaller environments introducing a SAN actually added complexity and reliability problems to an Exchange implementation. To counter this "SAN is a requirement" argument, Microsoft developed many of its best practices from the viewpoint that organizations did not need to deploy a SAN in order to support Exchange and framed many of its recommendations on how to support Exchange using DAS.

The trick is that as you move from small to large organizations, Microsoft's recommendations for best practices for Exchange storage configurations do not change to account for the storage technologies that these size organizations use. Rather they are tailored for the lowest common denominator such as what you might encounter in a small organization. So when large organizations roll out Exchange and call in an Exchange expert to help them configure and implement it, the expert may view storage from this more limited DAS viewpoint and does not recognize new features that next generation storage systems offer.

A prime example of this is the Windows 'diskpart' command. Exchange best practices call for administrators to run the 'diskpart' utility to align a partition with the underlying physical disk. If you do not run this command from Windows, Exchange data may be written to two disks instead of one because of how Windows, by default, starts a partition on the last sector of a disk. However by running 'diskpart' before installing Exchange, you can avoid unnecessary disk or stripe crossings since 'diskpart' starts a partition on the first sector of a disk so there is an increased likelihood that all Exchange data ends up on a single disk drive which improves Exchange performance.

Unfortunately 'diskpart' assumes that it is working with a traditional RAID architecture such as RAID 1 or RAID 5 so it follows traditional RAID boundaries. This is not the case with storage systems like the 3PAR InServ Storage Server which uses wide striping as it by default disperses Exchange data across most of the disk drives in its system. So while putting Exchange data on wide striping almost always results in better performance than even what following Microsoft's best practices will deliver, it may not technically meet the specifics cited Microsoft's best practices for implementing and configuring Exchange in these enterprise environments.

The good news is that 3PAR provides a mechanism to meet Microsoft's best practices for Exchange deployments and still automate storage management using wide striping without the need to re-introduce the overhead typically associated with managing and configuring storage volumes using its template feature. I'll get into how 3PAR's template feature works in an upcoming blog so that users can follow Microsoft best practices for Exchange storage configurations, still leverage 3PAR's wide striping feature and even potentially further improve Exchange's performance running on a 3PAR system.

Microsoft's Recommendations for Exchange Storage Configurations aren't Wrong, They are Just Out of Date

2009-05-21T10:00:00Z

No technology has become more ubiquitous or more critical to day-to-day business processes than email with Microsoft Exchange Server now arguably the email software that most businesses rely on. But even as Microsoft Exchange has become so widely adopted (65% of all organizations now use Exchange according to a recent report from Ferris Research) and the release of Exchange 2010 looming, information about how to optimally configure specific resources that Exchange uses still remains in short supply. Nowhere is this dearth of information more evident than in what new options are available to administrators as they look to configure and optimize the back end storage assigned to Exchange.

This is not to imply that Microsoft does not provide any guidance on how administrators should configure the storage that they allocate to Microsoft Exchange - Microsoft does do that. However the advice it dispenses are guidelines at best and it does not take into account new storage technologies that are now readily available to users. For instance, here is just one excerpt I pulled from the Microsoft Exchange Team's blog on how to configure, validate and monitor your Exchange 2007 storage:

"Mailbox Size/Mailbox Count... For example, if you have 4000 users on a server with a 250MB mailbox quota, then you need at least 1TB of disk space. Moreover, there are additional components which must be factored into the equation. If a hard limit is not set on the mailbox quota, it is difficult to estimate the how much capacity you will need."

On the surface, the recommendation seems pretty solid. But what that recommendation (and many of the others made in that blog) does not take into account is recent advances that have occurred in storage system technologies. This example of reserving 1 TB of storage for 4000 users so they have sufficient capacity for their mailboxes with 250 MB quota may not sound like a lot of storage but there are a number of problems with that recommendation:

First, that blog entry was written in 2007, not 2009. In 2009, mailbox quotes are closer to 1 GB than 250 MB (which is confirmed by another post on Microsoft's website) so the capacity requirements for users are now fourfold of what they were just 2 years ago
Second, only a fraction of the 4000 users (say 20%) are actually going to approach their 1 GB quota while the other 80% may only use a few hundred MBs of storage. Since an administrator may not know which 20% are going to need the 1 GB of storage, the administrator treats everyone the same and gives them 1 GB of storage. As a result, 2 - 3 TB of storage capacity is unutilized and wasted.
Finally, 2 - 3 TBs of wasted storage may sound like a rounding error in today's era of 1 TB disk drives but most Exchange implementations require that the data is stripped across multiple high performance disk drives to meet Exchange performance and availability requirements. So the amount of disk storage space and corresponding cost required to deliver on just the user mailbox requirements for Exchange may far exceed the minimum 4 TB of useable capacity that this specific implementation of Exchange calls for.

A better approach to thinking about how to configure one's storage for Exchange is to think about the new technologies that have become more widely adopted since 2007 like thin provisioning or wide striping such as what 3PAR's InServ Storage models offer. These features can dramatically reduce storage requirements, improve performance and enable organizations to do faster backups and recoveries of their Exchange environments. In a forthcoming blog entry, I'll elaborate at how new storage features like what 3PAR supports should prompt organizations to rethink the layout of the back end for their Exchange storage deployments and what new best practices they should consider when deploying Exchange with this next generation of storage systems.

Priming Oracle Data Files before Migrating to Thinly Provisioned Storage

2009-04-24T10:00:00Z

If there was ever a Golden Rule within database administration I'd venture to say that it would be: "Don't ever let your database run out of disk space". Doing so is a sure fire way to render a database inoperable, increase costs for staff fire-fighting, and, if you're the administrator, lose your job. For these reasons, DBAs and storage administrators have ventured on the side of extreme caution and over provisioned storage for applications--rendering valuable storage unavailable for other systems while inflating capital costs.

Data centers venturing to remedy improper use of storage have turned to thin provisioning as a mechanism to easily allocate space on a just-enough and just-in-time basis. Under thin provisioning, all applications can make use of spare storage, simplify storage administration, and reduce capital budgets.

But because of the internal nature of databases, migrating to thin provisioning can pose some interesting dilemmas if the free space is not dealt with before or during migration--resulting in a thin provisioning system that contains the same wasted space and consumes as much storage as the original system.

By their very nature, databases tend to leave empty space within data blocks through the improper setting of parameters such as block size, PCTFREE, and PCTUSED (in Oracle).Normal application processing such as INSERT, UPDATE, and DELETE operations that chain rows or exasperate the amount of free space within data blocks can also leave empty spaces. While migration and copy utilities for thin provisioning are good at removing free space within data files, some are unable to penetrate and see all free space within Oracle data files. For this reason it is a good idea to prep Oracle data files before hand to reclaim as much free space as possible.

What needs to happen is the compression of data and elimination of free space within the data blocks--basically a reorganization of each and every database structure that consumes storage. If you have an Oracle database, Oracle (10g and above) supplies a nice tool to perform in-place reorganization of data while keeping database objects online and available for application processing.

The Oracle SHRINK (ALTER TABLE...SHRINK SPACE) command will compact space used in a segment and return unused segments back to the table space which effectively packs data blocks as efficiently as possible. It is this shrinking of segments that makes it easier for the migration of data files to thinly provisioned volumes while completely removing wasted free space in them.

Recognize that the SHRINK command only compresses database objects and does not compress within table spaces or data files. As such, it may be beneficial, depending on how table spaces are being managed within Oracle, to look at the ALTER TABLESPACE...COALESCE and ALTER DATABASE DATAFILE...RESIZE commands for table spaces and data files respectively to eliminate all free space within Oracle data files before a migration.

As a member of the Oracle Partner Network and Oracle Unbreakable Linux Program, 3PAR and Oracle work closely together to bring a solution for managing complex database environments while maintaining optimal performance and controlling the increase cost of database storage. Powerful joint solutions such as combining Oracle's 11g database with ASM and 3PAR InServ Storage Servers with Thin Provisioning not only simplifies storage management but enables resource optimization through automated storage provisioning--resulting in a 30% reduction in storage related tasks and a projected 65% increase in productivity for database administrators.

Thin provisioning in a database environment is critical to the proper allocation and usage of storage. Working with a leader in thin provisioning and then properly evaluating your environment requires skill not only from vendors but within each company's data center as well. 3PAR's commitment to research, development, performance, and taking thin provisioning to new limits is well documented and can take care of the vendor side. Just remember, if a database has been improperly configured and maintained there may be additional steps to perform before migration. Shrinking, coalescing, and resizing Oracle data files should be considered the first step in preparing a database for migration to thin provisioning.

New 3PAR F-Class Midrange Storage System Skews Towards Solving the Problems of Enterprise Storage Shops

2009-04-06T12:00:00Z

Every time I run across an article that evaluates and ranks the latest midrange storage systems I always read it with a bit of a jaundiced eye. The latest lead article in the online March 2009 issue of Storage magazine is no exception. The article admirably tries to compare multiple midrange systems in an attempt to help users make an informed buying decision as to products that have proven their quality and reliability in actual use. Yet this article falls down in one important respect - it fails to take into account specific criteria for midrange storage systems that some enterprise storage shops will have.

Having sat in the decision-maker's seat before, I did not see mentioned in this report some of the criteria discussed that often influenced my buying decisions of midrange systems. Specific features that were becoming more important to me were how or if the midrange storage system would interface with a Tier 1 storage system or if the features that a midrange storage system possessed would give me the confidence to buy and use it in lieu of a Tier 1 storage system.

Of course, not every user or shop has the same criteria that I had for the purchase of midrange storage systems. In fact, this Storage magazine article plainly states that its results were skewed towards small and medium-sized enterprises (under $1 billion in revenue). But an organization that generates billions of dollars in revenue, supports mission-critical applications, or manages hundreds of TBs of data on different tiers of storage selects midrange storage systems based on a different set of criteria than what SMB shops might use.

Large organizations increasingly recognize that all storage systems that they bring into their environment need to be helping them build towards creating a more manageable and scalable storage environment. Specific key features that the storage system managers in these enterprise shops should now be looking for in midrange storage systems include:

Interoperability with Tier 1 storage systems. Enterprise organizations are looking to making their environments easier to manage, not more complex. So while midrange systems are often less expensive than Tier 1 storage systems, they may force their administrators to learn new management interfaces, purchase new software to manage them and ultimately make their more environment more complex. Conversely, if the midrange system offers the exact same firmware, the same software features (thin provisioning, replication, snapshots, etc.) and management software, the management picture changes. Now organizations can manage them in the same way as, and even exchange data with, Tier 1 systems so the lower price of midrange systems looks even more appealing since organizations can manage them as part of their overall storage infrastructure.
The same high availability, reliability and performance characteristics as Tier 1. Most of the time this is where midrange storage systems come up short. They can scale to ridiculous levels of capacity but then only offer 99.9% of uptime or only available in Active-Passive or Dual Active controller configurations. Even if a midrange storage system is available in an Active-Active controller configuration, the back-end LUN is still active on just one controller so there is no means to scale performance out to that configuration. Enterprise organizations sometimes need to opt for Tier 1 systems for these reasons simply because the application demands it and Tier 1 systems are the primary means to deliver it.
A smooth upgrade path into a Tier 1 environment. This is where most midrange systems fall down. Applications rarely stay the same and yesterday's departmental application becomes today's mission critical application that is rapidly growing and requires constant uptime, more capacity and more performance. This often requires a painful data migration to move that application's data to a storage platform where it meets that application's new requirements.

This is really the set of problems that I see the new 3PAR F-Class midrange storage systems that was announced today best addressing. Yes, it can compete against any of the traditional midrange storage systems and probably do so in quite a favorable manner. But where it really outshines its competition is in its ability to let enterprise shops nestle it in nicely next to their existing 3PAR T-Class Tier 1 storage systems.

The F-Class offers availability, reliability and performance features that are comparable to the T-Class though these features are, of course, scaled down since the F-Class is a midrange storage system. But what makes the F-Class notable is that it possesses the same software as the T-Class while the upper end F400 model delivers the same hardware architecture as the T-Class.

This combination of software and hardware features is important in two key ways. First, all of the T-Class' software features are available on both models of the F-Class so organizations can manage both F-Class and T-Class in the same manner to which they are accustomed now on the T-Class. In fact, they can even move data between these different models without requiring 3rd party replication tools.

Second, the T-Class' mesh-active controller architecture (this is 3PAR's new name for its controller architecture so if the name looks new, that's why) also found its way into the F400. This feature is truly a high-end feature as it makes each volume that is presented to servers active on each of the up to four controllers that the F400 supports. The Active-Active controller feature is one that you rarely find on midrange arrays with two controllers, much less four, so this is a radical departure (in a good way) from what other midrange systems can offer.

Many enterprise organizations are itching to bring midrange storage systems into their environment that they can manage as part of their broader Tier 1 storage infrastructure. The F-Class from 3PAR answers the demands of these organizations. Yes, SMB shops will likely still be attracted to the Tier 1 features and management characteristics that the F-Class offers and for these size shops looking for this type of solution, it certainly fits the bill. But when one looks at where the F-Class best fits and what problems it was designed to address, it certainly seems to skew towards solving the problems that enterprise storage shops possess.

Thin Provisioning - One of the Silver Linings in Today's Environment of Doom and Gloom

2009-03-25T10:00:00Z

As the global economy continues to weaken, one might begin to think that all technology is suffering along with it. In fact, a number of recent press releases from Gartner do little but reinforce the doom and gloom:
Worldwide Server Shipments and Revenue Experience Double-Digit Declines in Fourth Quarter of 2008
Gartner Says PC Industry Will Suffer Sharpest Unit Decline in History in 2009
Gartner Says Worldwide Semiconductor Revenue On Pace to Decline 24 Percent in 2009
Gartner Says Worldwide Vertical Market IT Spending Will Be Flat in 2009

Yet in the face of all this, the storage market continues to enjoy steady growth. Even Gartner highlighted that the storage industry is still growing citing 11.3 percent worldwide growth in the external controller-based disk storage market in 2008. Another analyst, Jon Collins with Freeform Dynamics, states, "There is no recession in storage for many reasons, but the two main ones are: the continuing growth of rich media, and compliance. As long as we need to keep records of everything, the need for more storage will always be with us."

But as DCIG has been writing about in this blog for some time now, we must remember that the purchasing of storage just because one needs more capacity is futile. While the economy is taking a dive, storage technology as a whole has the potential to prosper and help companies beat tightening budgets. When the economy goes south, there tends to be a greater necessity and, as a result, a greater push for companies to find innovative technologies that will help them reduce their TCO. DCIG's blog archives already cover the importance of using innovative and solid technologies that help organizations of all types and why it is critical to their survival. Previous DCIG blogs on this topic include:
Troubled Economic Times Call for Solid Technology Answers
The Greenback might be Floundering but Green IT is Proving Profitable
CSUEB Sees More Green with Less Hardware, Lower Power Requirements and PG&E Rebates
Even Nonprofits Need to Innovate During Tough Economic Times
Dynamic, Agile Infrastructures are Key to Picking Today's MSPs

Interesting to note, back in 2007 Gartner stated that Thin Provisioning Is Revolutionizing Storage Management and that "insatiable storage growth, coupled with staff shortages and budget constraints, will make thin provisioning a must have feature"--increasing utilization rates, staff productivity, and performance while reducing footprint and TCO. Fast forward to today and we see storage provider 3PAR leading the way with their 3PAR InServ T-Class Storage Servers with the Thin Built In ASIC and 3PAR Thin Provisioning software.

Just as the economy is floundering, 3PAR is seeing an increased demand for its highly virtualized, cost- and energy-efficient storage arrays. And in the face of corporate failures and increasing numbers of layoffs, 3PAR is expanding its operations by adding to its European team and opening new offices including a second London location to support the company's regional headquarters--beating the acquisition, merger, or failures seen by other storage vendors that have not been able to deliver on innovation. But by offering customers a highly cost-efficient alternative to traditional storage that enables them to lower both CAPEX and OPEX (of huge importance these days), 3PAR has been able to sustain growth by having products that appeal to both the IT administrator and the budget holder and enabling them to do more with less--a critical selling point in today's tough economic times.

Utility Storage is the Right "Technology Inside" For Flailing Web 2.0 Storage Infrastructures

2009-03-16T10:00:00Z

The shift from Web 1.0 to Web 2.0 is a result of both a shift in technology and a shift in behavior. Technology has come a long way from the static pages, frames, and one way flows of information seen in Web 1.0. Web 2.0, with its rich internet applications (RIA), content syndication, and APIs have decentralized website content and created an atmosphere where users have become both consumer and contributor of information. Static is out and dynamic content generation with rich backend databases for 1.5 billion internet users to interact is the norm.

The importance of data in Web 2.0 was stated best when Tim O'Reilly said, in the article What is Web 2.0, "Data is the Next Intel Inside". Data is what drives Web 2.0--offering users the ability to interactively and collectively use and manage information and applications across the Net. Content and applications are alive as mashups aggregate content and applications from multiple websites and add to a collective value for users.

But with all this great dynamic Web 2.0 experience comes the high potential for application and content delivery performance problems. Accepting and delivering content from users, to be shared amongst many users, places a high level of stress on backend systems. When this happens response times will suffer and users will quickly propagate to other more responsive sites.

In the article The Coming Web 2.0 Train Wreck, Jim Metzler states that while both client side and server side hardware impact Web 2.0 performance and response times, it "is the server side delay that is the critical performance bottleneck that has to be addressed in order for Web 2.0 applications to perform well". The reason being that, "unlike clients, servers suffer from scalability issues. In particular, servers have to support multiple users and each concurrent user consumes some amount of server resources: CPU, memory and I/O".

Storage is a key component for Web 2.0 applications. Websites such as LifeScript.com, which has 4 million unique visitors, 30-40 million page views a month, and 10 million newsletter subscribers, relies heavily on a storage infrastructure that can store and deliver dynamic content to their readers. It needs reliable and highly available content delivery to drive traffic to its website or LifeScript's business-to-consumer model would deteriorate.

In response to this direct need to grow its storage infrastructure and provide for easy and reliable growth without risking performance levels during peak usage periods, LifeScript purchased the 3PAR InServ Storage Server with 3PAR Dynamic Optimization for non-disruptive and autonomic system resource balancing and 3PAR Thin Provisioning software for capacity savings.

Using 3PAR, LifeScript's Web 2.0 portal and email processing engine now supports three billion transactions per month at a third of the cost than its previous storage system. LifeScript's new infrastructure has also reduced average time to provision new servers from several hours to just minutes. Jack Hogan, Chief Operations Officer for LifeScript, says, "3PAR offered us both capacity savings and intelligent integration with VMware to curb storage costs and provide greater performance and agility."

Success of Web 2.0 websites are directly proportional to how well customers interact with information. The explosive growth in Web 2.0 is forcing companies to seriously re-examine their ability to keep pace with the high demand for dynamic content. Inflexible and inefficient traditional storage systems need to be upgraded - and in many circumstances replaced - to enable the safe allocation of upfront logical capacity while delivering high performance. Using 3PAR's native tiered-storage capabilities and its Thin Provisioning software, companies can now dynamically deliver on the new and unanticipated capacity, performance and scaling demands of Web 2.0.