Just about a year ago 3PAR announced an innovative blueprint for the virtual datacenter to help companies meet the increased demands on IT. Its objective was to eliminate the need to dedicate physical computing and storage resources to specific applications that have led to spiraling server and storage costs, datacenter sprawl and time-intensive provisioning and change management. Coined 3cV for the combination of the 3PAR InServ Storage Server, HP c-Class BladeSystem and VMware Infrastructure, it forms a complete utility computing offering that enables customers to halve server, storage, and operational costs and gain instant agility.
An early believer in this program was Rich Jackson, Vice President of Technology Operations at Hilton Grand Vacations. At that time, he made the following statements:
Fast forward to today and Hilton Grand Vacations was recently awarded first place by the Technology Managers Forum in the category of Storage Management as part of their 2008 Best Practice Awards program. As a professional association for IT managers at mid-to-large corporations, one of its missions is to recognize achievements in technology and how companies use it for a competitive advantage. Some of the benefits experienced by Hilton Grand Vacations include:
Hilton Grand Vacations' benefits cross boundaries of physical storage manageability, operational efficiency, application performance and outside services. This is a direct result of the 3cV blueprint which directly attacks IT, server and storage consolidation to provide a highly virtualized utility computing. This new virtual environment delivers exceptional levels of agility and efficiency in drastically changing IT environments. Building this form of flexible and highly virtualized utility computing environment can actually be simple when there is a proven blueprint available. Because all components of the 3cV blueprint are qualified to work together, companies can quickly deploy a solution in response to virtualization initiatives that will provide exceptional levels of agility, performance, and efficiency.
]]>But just as the dot-com bomb helped us realize faulty business practices, the possibility of recession will more than likely help us realize how to better spend corporate monies. Gartner notes that the impact on IT budgets in 2009 will not be as extreme as those suffered in 2001; though it only projects a 2.3 percent increase in 2009 versus its previous 2009 projections which pegged increases at 5.8 percent. Since IT budgets for 2009 are certain to remain uncertain in the coming months, this is obviously the time to rethink future direction and how to optimize current IT expenditures.
The storage industry and how companies utilize storage is ear-marked as one sector that will continue to do well despite a rocky economy. Companies continue to need storage as the amount of data that companies manage continues to grow at an unabated rate. But this does not mean companies can unabashedly throw more storage at their storage problems. Instead companies need to become more conscientious about the quantity, type and value of the storage that they acquire so they need to look for innovative products and features that will allow them to do more with less. Innovative technologies such as cloud computing, SaaS, utility computing, managed services and thin provisioning will now all play key roles as these are pay-as-you-go services, have lower up-front costs, enable IT to enable to respond more quickly and add new features without needing to invest in new hardware.
In this respect, 3PAR is one company that is leading the charge by providing utility storage and thin provisioning features. 3PAR offers IT management a "utility" storage array platform that reduces the cost, complexity, and risks associated with deploying and managing a storage infrastructure. They provide a strong foundation though its Thin Provisioning feature so companies can provision virtual capacity to applications once and purchase pooled physical capacity only as applications truly require it for written data, eliminating the over-purchasing and over-allocating of storage resources. And, with 3PAR's recent announcement on October 16 to integrate its thin provisioning technology with Symantec's file system-level information to automatically reclaim unused space within thinly provisioned virtual volumes, we can expect continued optimization of storage utilization coupled with advanced infrastructure automation.
3PAR's Utility Storage further helps companies reduce capital expenditures by enabling them to initiate larger numbers of server consolidation projects, deploying more virtual machines (VMs) using fewer resources. With the use of 3PAR's Thin Provisioning, organizations can provision large amounts of virtual capacity up-front to the ESX server, and then only purchase physical capacity only as additional VMs or provisioned or application data is written. This eliminates allocated but unused capacity while simplifying storage administration. The combination of Thin Provisioning and VMware is 'do more with less' on steroids - it should not to be ignored in times of budget stress.
In these volatile times, 3PAR isn't hiding under the rock. Utility storage, and its impact on utility computing, is no longer just being promoted by a broad range of industry analysts, but it is enabling IT to shift strategically from meager store-houses of information to leveraging innovative technologies to provide corporate advantage even during an uncertain economy. Clearly 3PAR has come a long way, has broken out of the start-up mode, and is advancing storage technology to new levels--just in time for IT to breath a sigh of relief during these troubling times.
]]>There is good reason for this. They want to design a system architecture that will provide acceptable levels of service and performance today as well as in the future. Unfortunately there are opposing forces such as service level agreements (SLAs) and operating costs within organizations with which they must contend. For example:
Making matters worse, many administrators and architects find it difficult, if not impossible, to agree upon underlying system configurations based on application and service level mix. The simple selection of a RAID level can become a huge design battle. On numerous occasions as an architect and administrator I've had heated discussions about the different RAID levels that are better suited for some storage structures than for others. But because application mix, service levels and future requirements were unknown, IT was often forced and locked into architecting a storage solution that left many questions unanswered.
Because there are often many unanswered questions about current and future application mixes and their future impact on performance and service levels, I've become quite excited about 3PAR's dynamic optimization feature and its ability to deal with the increasing unpredictability of cloud computing and next generation data center environments. Using simple storage-side commands, administrators can alter the underlying storage structure in a non-disruptive fashion.
Also through dynamic optimization, IT can easily change, convert, or adjust key storage features such as RAID type, degree of utilization for processors, drives, and ports, the placement of data on disk, and the type of drives being used (FC or SATA-class drives). In effect, IT can alter the cost structure and service level attributes of an application volume rapidly, non-disruptively and on-demand. Combine 3PAR's dynamic optimization with its utility storage and you circumvent the two previously mentioned opposing forces:
Planning for the future is tough as I can testify based on the numerous occasions I've been brought into organizations to look into the reasons behind faulty performance of applications and databases and then figure out how to fix them. This is particularly true when considering virtual server technologies and next generation data center infrastructure. While there are many instances where application and database performance tuning can bring service levels back into compliance, it has been my experience that there are even more application, database, and service level problems that can only be solved from a hardware perspective.
The beauty of a 3PAR solution is that it doesn't lock you into a particular configuration and provides the flexibility to deal with unpredictable application requirements cost-effectively. Using 3PAR, organizations can now align application and business requirements with data service levels easily, precisely, and on demand -- achieving optimal performance and service levels while lowering operational costs.
]]>Now the type of infrastructure needed for that concept has emerged and it even has a name: cloud computing. While sites like Wikipedia can provide a more detailed explanation of what cloud computing means, at a high level cloud computing is all about enabling companies to dynamically provision computing power, network bandwidth and storage resources to their applications when and where they need it. In effect it is enterprise IT delivered as a service via the internet, typically by external B2B or B2C service providers. Effectively, cloud computing is Software-as-a-Service (SaaS) or Infrastructure-as-a-Service (IaaS) delivered from a hyperlink. It is this internet-based delivery system that distinguishes the approach and gives cloud computing its name.
Using cloud computing companies can increase or decrease the amount of resources that they dedicate to specific applications while incurring the expense or realizing the savings at the time the change is made. Granted, this is a bit of a simplistic explanation of how cloud computing works. But for cloud computing to even come to the point where companies can contemplate using it for some of their application needs, a lot of other technologies had to mature along the way. However one of the promises of cloud computing is that it promises to convert enterprise IT from the fixed cost model that exists today to a new variable cost model.
To effectively support cloud computing, the underlying infrastructure needs to support a couple of key characteristics. First, these systems are autonomic; that is, built-in automation technology eliminates manual planning and management tasks and allows the platform itself to respond intelligently to application needs. If the system requires human interaction to allocate and manage resources, it's not a cloud: it's just a traditional (distributed computing) data center.
Second, cloud computing infrastructure is agile, able to respond immediately to spikes in demand or changing workloads. In other words, built-in virtualization coupled with clustering technologies allow for rapid changes in growth or service level requirements. If the system requires hours, days or even weeks to react to new application or user needs, it's not a cloud: again it's just a traditional (distributed computing) data center.
Of these technologies, probably none is more important to delivering on the reality of cloud computing than virtualization. VMware went a long way towards popularizing and educating people on the concept of internal server virtualization. But server virtualization in whatever form it takes (Citrix's XenServer, Microsoft's Hyper-V, etc.) is only one piece of the total virtualization puzzle. To realize the full potential of cloud computing, it needs all of the virtualization pieces from the server through the network down to the storage system to interoperate with one another to deliver the autonomic, agile infrastructure that a dynamic cloud computing environment demands.
Without a doubt, automation, clustering and internal virtualization technologies absolutely critical to delivering on the reality of cloud computing. For companies looking to build cloud computing infrastructures or outsource their applications to hosted service providers that utilize cloud computing, these technologies must exist at the core of the server and storage infrastructure for the promise of cloud computing to actually be realized.
Not only do these cloud computing environments eliminate manual tasks like provisioning and data migrations, but the systems they use are aware of application requirements and dynamically respond to their needs. For instance, storage platforms like 3PAR's InServ Storage Servers are already used by hosted service providers and Software-as-a-Service companies that offer cloud computing. The InServ Storage Server combines automation, clustering and internal virtualization within all of its storage. It only autonomically allocates the storage that applications need, and dynamically load balances and re-allocates storage resources based on the changing requirements of the business. In so doing, it addresses the need of the customer to rapidly manage unpredictability within their infrastructure, which enables them to serve their clients better, faster while still turning a profit.
Cloud computing is still in its early stages but with recent advances in virtualization at all levels of the infrastructure and a broader understanding of its benefits, cloud computing is moving off of drawing boards and into real world implementations. As it does so, it is putting all corporate IT on notice. They cannot expect to cost-effectively and efficiently deliver on tomorrow's needs using yesterday's technologies. Products like 3PAR's InServ Storage Server are prime examples of the type of products that will make up tomorrow's virtualized infrastructures within these emerging cloud computing environments.
]]>The Linux operating system has long been a staple within IT infrastructures to accomplish these goals as it enables IT departments to utilize low-cost commodity server hardware to power enterprise class software. Taking advantage of low-cost servers to run Linux is one of the differentiating factors when organizations attack the total overall cost reduction puzzle for servers. Robust features, such as the clustering of Linux servers to replace larger, more expensive servers, while providing for higher levels of flexibility, availability, and scalability, only further drives home the significant price/performance benefits that prompts companies to adopt this approach.
While many benefits are realized with the deployment of key features such as clustering and low cost server hardware, there is the potential to increase complexity and administrative costs. Trying to hire a complete and competent workforce to manage Linux servers, or turning to outside consultants to deploy, manage, and maintain this new un-charted IT infrastructure, can be risky and introduce new costs. Simplifying the support model by utilizing outside services is critical to reducing these administrative burdens and costs.
Organizations can now take advantage of enterprise level support from companies such as Oracle that have a high stake in the success of Linux. Under the Oracle Unbreakable Linux Support program, Oracle offers true enterprise-class support for Linux with premier services such as free binaries, patches, fixes, updates, backports, and Oracle global support, all with the intent of helping organizations succeed with Linux at a significantly lower cost.
The Oracle Unbreakable Linux Support program offers three levels of Linux support so an organization can easily choose a support level that caters to their individual needs and comfort levels with Linux. As a worldwide enterprise support provider, Oracle has a team of over 7000 support professionals including dedicated Linux engineers to take part in complete testing and optimization initiatives as well as expert installation, configuration and Linux stack deployment so Oracle is well positioned to offer superior support services.
Also, as part of the Unbreakable Linux Support program, Oracle performs comprehensive testing and optimization of Linux with storage vendors. This ensures: compatibility between Linux and storage solutions; the promotion of storage solutions that provide performance and innovation; reductions in system complexity and costs for enterprise solutions; and, out-of-the-box readiness for immediate system deployments.
As part of this initiative to help simplify management and reduce costs, 3PAR recently joined the Oracle Unbreakable Linux program to extend this out-of-the-box readiness to encompass its next-generation storage. 3PAR's Utility Storage lets you provision virtual capacity to applications once and purchase pooled physical capacity only as applications truly require it for written data, so companies can provision or meter storage as they need it. This extends the basic, low cost tenets of Linux to the storage systems that companies use with their Linux clusters by enabling companies to simplify the provisioning of storage during periods of peak demand while eliminating purchasing and deployment delays.
However do not confuse the management simplicity of 3PAR's Utility Storage technology with other storage systems. The 3PAR storage solution in both OLTP and OLAP environments, as the recent SPC-1 benchmark shows, delivers not only supply high performance but also scales as Oracle workloads increase. Adding to the utility of 3PAR storage, the fact that workloads are striped across all array resources deliver robust and load balanced performance, eliminating the administrative nightmares associated with application performance tuning.
The 3PAR and Oracle partnership in the Oracle Unbreakable Linux program provides a new framework for organizations to simplify system and storage management, consolidate support and offer cost savings from the storage infrastructure level up to the application environment. Complete dedication from the community is what now makes Linux a viable solution for many organizations now. 3PAR's added support for the Oracle Unbreakable Linux program is one more way that companies can now lower their total costs of data management while simplifying the delivery of enterprise IT as a utility service.
]]>If you followed that little rhyme you surely picked up on the fact that previously the only criteria that mattered for prioritizing how you spent your money was on storage capacity. The truth of the matter is that many organizations toil over how to efficiently pack-in data into the smallest storage unit. Then when things go Pop!, go south, or grind to a halt storage administrators, architects, programmers, DBAs, and IT management are called into action to scramble and find out why, all of a sudden, the user experience or the application run time has degraded.
Many times Jerome or I have been called to investigate and fix performance issues. More often than not, storage was purchased based solely on the reason of capacity. For instance, disk storage requirements for a new database were calculated at 100GB so 100GB of inexpensive disk storage was purchased--not understanding that database attributes such as transaction rates and throughput should also be considered in the purchase.
For this reason, IT's inability to properly purchase a storage solution that delivers both capacity and speed, we got excited by the prospects of this morning's announcement of the new T-Class storage servers from 3PAR. Uniquely combining utility storage with exceptional performance is a rare combination. The facts are storage administrator talent is becoming sparse, often inadequate, and is expensive to cultivate in-house. Deploying utility storage from 3PAR helps eliminate many of the costs associated with developing the talents of storage administrators in-house or using high-price consultant. Using 3PAR's new T-Class storage servers, continually tuning storage performance and allocating space should become a task of the past.
Proof in point, 3PAR's new T-Class storage servers recently set a new Storage Performance Council SPC-1 performance record for single-system storage arrays--doubling the performance of previous generation 3PAR S-Class arrays. These T-Class arrays, specifically the T800, generated nearly 225K IOPS in a published SPC-1 result. More significant, this array was maxed out to 83% capacity utilization (77,824 GB in a mirrored configuration) and was taken straight out of the box without using any complex configurations, performance tuning, or employing high-tech performance tricks, such as short stroking, to pump up performance.
Built on third-generation architecture, 3PAR is the first storage vendor to incorporate efficient silicon-based thin technologies into system hardware. The 3PAR T-Class arrays capitalize on a new, innovative Thin Built In architecture to introduce its new zero-detection functionality. Zero-detection finds unused capacity within "fat" data volumes with the ability to provide fat-to-thin volume conversions.
Fat-to-thin volume conversions are critical for those companies looking to migrate to 3PAR T-Class storage servers from existing storage systems. Zero-detection identifies over-provisioned volumes by looking for long strings of zeros in LUNs or volumes being migrated to the T-Class storage server as zeros are indicative of unallocated space on incoming LUNs.
By stripping out these zeros, companies can turn previously over-provisioned LUNs on older storage systems into thinly provisioned LUNs on the 3PAR T-Class storage system. This technique boosts capacity utilization by removing allocated but unused space from volumes. The T-Class storage servers implemented this within an application-specific integrated circuit (ASIC) instead of a software solution to ensure that it will not steal valuable CPU and memory resources not negatively impact application performance.
3PAR's new zero-detection feature addresses a long-standing criticism of storage systems about thinly provisioned volumes - how to easily migrate existing traditionally-provisioned, fat LUNs and convert them to thinly provisioned LUNs. 3PAR's implementation of zero-detection at the block level using its purpose-built ASIC now offers IT Management storage arrays that reduce the cost, complexity, and risk of deploying and managing a storage infrastructure. In so doing, companies that implement 3PAR T-Class family of storage systems should get the Pop! in storage capacity and performance without needing similar Pops! in administrative and capital costs to realize the inherent value that these storage systems provide.
]]>End-users in distributed environments expect to retain the same privileges of self-service storage (freedom to make changes, provision storage and manage applications) as they did prior to the storage consolidation. However end-users can loose this flexibility and freedom to manage storage themselves unless they pick a storage system that is specifically designed with the internal virtualization capabilities to accommodate these expectations. To do so, companies need to look beyond storage providers that initially designed their storage systems for mainframe storage consolidations and later retrofit them for distributed, open system storage consolidations. Instead they need to select systems that are intended more specifically for virtualization-based storage consolidations.
In these circumstances, companies may be better served by considering a next-generation storage system like the 3PAR InServ Storage Server. It mirrors other enterprise storage systems in that it provides large numbers and different types of storage connections (iSCSI - 32 ports, FC - 128 ports), processing power (up to 8 controller nodes) and large amounts and varying tiers of storage capacity (up to 600 TB). Where 3PAR distinguishes itself from other enterprise storage systems is through the Virtual Domain feature found on its InServ Storage Servers.
Virtual Domains preserve the storage management freedom that departments and business units were accustomed to prior to consolidation of their application data on shared storage systems. It resembles a similar concept in server virtualization, the virtual machine, in that each Virtual Domain is administratively secure and appears to the user as a dedicated (virtual) array. However each Virtual Domain leverages the massively parallel resources of the InServ, delivering higher performance per domain than would otherwise be available on a dedicated system. Like virtual server environments, each Virtual Domain operates independently of other Virtual Domains so after the consolidation is complete, individual departments or business units can be assigned one or more Virtual Domains which they then have full authority to manage as they best see fit.
Each Virtual Domain offers all of the storage management functions that the underlying 3PAR InServ Storage System offers, including thin provisioning, snapshots and the creation of storage volumes with guaranteed service levels. Capacity thresholds can also be assigned to each Virtual Domain by storage tier. However since each Virtual Domain is its own separate, logical entity, if an Virtual Domain administrator makes a configuration change that results in an error, only servers attached to that Virtual Domain are potentially impacted. This differs from other consolidation environments where any configuration mistake could ripple out to impact any server attached to the storage system.
Virtual Domains should also not be confused with hard partitioning which is employed on some systems. Hard partitioning divides and dedicates specific hardware resources (controllers, memory, storage, port connectors) into partitions within the storage system. The challenge with this approach is that once the partitions are set, the resources are pre-dedicated to that partition and utilization suffers - particularly where performance is a consideration and large numbers of disks must be pre-dedicated to physical partitions to meet certain service levels.
Using Virtual Domains, all Virtual Domains on the storage system have access to all hardware resources and are distributed across them. This gives administrators the flexibility to configure resources within and across Virtual Domains as application demands fluctuate over time and prevents over- or under-allocation of resources which can occur when hard partitioning is employed.
Storage consolidation is a dream that many companies are seeking to realize but what they often find is that the reality of consolidation does not match its promise. The Virtual Domain feature on 3PAR's InServ Storage Systems changes that. It takes the concept of server virtualization that is already known and understood and applies that to its storage system. In so doing, corporate business units and departments can, post-consolidation, retain the storage management flexibility and freedom to which they are accustomed while companies can realize the technical and financial benefits of consolidation that they initially sought when they implemented it.
Part 1 in this series examines some of the hidden, intangible costs of consolidation once the consolidation process is under way.
]]>Despite the server sprawl that most companies are looking to deal with, a more insidious problem that they also recognize that they need to deal with is corporate desktop management. Companies may have thousands or even tens of thousands of desktops in-house and while the hardware costs for these desktops have become fairly nominal, the soft costs of configuring and supporting them mount. Add in the hidden costs of accessing and searching the data on these desktops should companies find themselves subject to an eDiscovery request and the costs of desktop management can quickly escalate to match or even exceed that of server management.
Yet the problems of implementing desktop virtualization are equally subtle. Desktop virtualization has never really achieved corporate adoption for one simple reason - a cost-effective, highly available, highly scalable back-end infrastructure. In this case, the nominal cost of desktops has worked against the adoption of technologies like VMWare's Virtual Desktop Infrastructure (VDI) because of the up-front investment in backend infrastructure required to make virtual desktops a reality.
The new Thin Copy Desktop feature on the 3PAR InServ Storage Server coupled with its existing Thin Provisioning and Virtual Copy features disrupt this mindset in a number of ways and make the wide-scale deployment of VMware VDI feasible for companies in the following ways:
Virtualizing the desktop infrastructure is an objective that many corporations have desired for years though limitations and deficiencies in the corporate infrastructure have precluded them from pursuing this objective. The combination of 3PAR's InServ Storage Server and VMware's VDI now makes such a reality possible and justifiable - both technically and economically. As it does, it changes the conversation around desktop virtualization from a "nice-to-have" feature that is often impractical to implement to making it a viable and realistic virtualization option.
]]>A major part of the problem stemmed from expectations set prior to consolidation. Departments, business units, application owners, etc. operated and managed their storage systems in a manner similar to how they managed their servers: In a distributed fashion. Managed this way, each department or business unit often had its own servers and storage so they had the freedom to manage it in whatever manner best suited them. Those units doing development, research or testing therefore could make frequent, ad hoc changes to their storage systems.
Conversely those business units or departments that managed systems in high impact (i.e. revenue generating, public exposure, etc.) environments could only make a system change after the change was analyzed from every possible angle. Even then, the change could only take place at certain times and days after everyone and their brother knew when it was going to occur.
The practical ramifications of consolidating storage from these different departments showed up only after the consolidation occurred. Suddenly business units and applications that needed rapid changes to their storage configurations were hamstrung by these new requirements. They needed to adhere to more stringent requirements of production systems since production applications often took priority over applications in test and development.
Even in situations where consolidations of similar production systems occurred, the question of who should take ownership in the management of the storage regularly arose. Should it be the one with the most experience in storage management? Should it be the one who has done it the longest? Should the different departments share the storage management responsibility but if they do, who is responsible should something go awry?
So essentially what is emerging out of this configuration is a new set of rules for storage systems that are implemented as part of storage consolidations in distributed environments so they can deliver self-storage storage. Specifically:
Companies now want it all: they want the flexibility of self service storage that they experienced while using distributed computing but they also now need and want the cost and power saving benefits that storage consolidations deliver. Meeting all of these requirements is something companies can only achieve using new storage systems such as 3PAR's InServ Storage Server. In part 2 of this series, I'll take a deeper look at how it delivers on these features and help companies not just implement storage consolidation but live the storage consolidation dream.
]]>