Essentials Guide CONSIDERATIONS FOR SELECTING ALL-FLASH STORAGE ARRAYS

Essentials Guide CONSIDERATIONS FOR SELECTING ALL-FLASH STORAGE ARRAYS

M ost storage vendors now offer all-flash storage arrays, and many modern organizations recognize the need for these highperformance systems, however deciding on which of the many all-flash arrays is best for your organization can be difficult. In this e-guide learn from storage expert George Crump criteria to help IT pros decide whether performance or function is most important when choosing all-flash storage arrays. Additionally, explore the various implementation types and the benefits and drawbacks of each. PAGE 2 OF 16

CONSIDERATIONS WHEN USING SERVER-SIDE CACHING IN VIRTUAL ENVIRONMENTS George Crump of Storage Switzerland outlines issues associated with using server-side server in this Expert Answer. ARE THERE SPECIFIC CONCERNS WITH USING SERVER-SIDE CACHING IN VIR- TUAL SERVER ENVIRONMENTS? Yes, there are several. First, how does server-side caching impact virtual machine migration? If a read-only server-side cache is being used with a shared storage array, there is limited risk of data loss. The challenge is when the virtual machine is migrated; its data needs to be re-qualified on the target server. Users will see hard drive performance until the target server can qualify that virtual machine s data. If the cache is a write cache, then the cache needs to be flushed to the shared storage system prior to the virtual machine migration. An increasing number of caching software suppliers are integrating with the hypervisors to manage through these issues. Many now integrate to acknowledge the migration event and take appropriate action, such as flushing PAGE 3 OF 16

the cache. In addition, some caching vendors can mirror the cache to a shared SSD on the network. During normal operation, reads are served from the cache in the server, but if there is a server-side cache failure or if the virtual machine is migrated, all reads can come from the shared SSD on the network. This is also an ideal configuration for caching writes, thanks to the redundancy the mirror provides. Some vendors are also providing the ability to move the cache metadata between systems. When a migration event for a VM is triggered, its cache metadata is transferred to the target server. While the data does need to be reloaded from the shared storage system, the target cache knows exactly which data to get; no data analysis needs to take place. PAGE 4 OF 16

ALL-FLASH STORAGE ARRAYS: PERFORMANCE VS. FUNCTION George Crump offers criteria to help IT pros decide whether performance or function is most important when choosing all-flash storage arrays. Nearly every storage vendor now offers all-flash storage arrays, and IT professionals are beginning to recognize the need for these high-performance storage systems. But how does an IT pro decide which of the many all-flash arrays are best suited for their organization and performance demands? PERFORMANCE VS. FUNCTION As the all-flash storage array market begins to mature, there are two categories of arrays emerging. The first are all-flash arrays that were designed from the ground up to be all-flash arrays. They typically have optimized hardware designs that focus on extracting the maximum possible performance from the flash within the array. The vendors in this space are almost all emerging technology companies or startups. In most cases, their focus on hardware and performance is at the PAGE 5 OF 16

expense of storage software services. These are the features that many storage administrators now count on to do their jobs, providing capabilities like snapshots, replication and cloning. These arrays are known for generating millions of IOPS per system. However, there really is no established method for how those high IOPS numbers are obtained. They can be generated from a single workload or multiple workloads accessing the system at the same time. The other category is made up of all-flash arrays that are more featureoriented. These are typically systems from established vendors, as well as a few startups, that choose to focus on the software functionality (providing a feature-rich experience), often at the expense of maximum performance. Typically, these systems either use legacy hardware from the established vendor and retrofit their old arrays with solid-state drives (SSD) or, in the case of a startup, use off-the-shelf hardware to keep costs down. These systems can often generate 200-400k IOPS per system. Some scaleout, software-rich systems will claim an aggregate performance of millions of IOPS as well but, as mentioned above, the devil is in the details. They typically have a performance limit per volume or per node within the scale-out cluster. This means they can scale to millions of IOPS like the performance-focused PAGE 6 OF 16

systems described above, but it takes many nodes to get there and to see that extreme performance requires multiple workloads all running concurrently. A scale-out system cannot deliver millions of IOPS to a single workload or thread. WHICH IS BEST? We are often asked which method is best. The answer, as usual, depends on the needs of the data center and the specific applications that are running. Most data centers, while performance-constrained, are not constrained to the point that they will typically exceed the baseline performance of a feature-rich allflash array. Also, most organizations will take great comfort in the availability of the feature sets they have become accustomed to from legacy hard disk arrays. There are with a need for more than a half million IOPS, but it s how those IOPS are needed that will help determine the best system for a particular data center. If the need for performance is distributed across more than a few workloads, the all-flash systems that can provide scale-out linear performance growth are ideal. If the environment has a single workload that needs more than half a million IOPS, then the performance-focused systems are needed. As stated above, these systems can provide millions of IOPS to a single workload. PAGE 7 OF 16

MIDDLE GROUND? Is there room in the middle? Does a storage system exist that can meet the needs of a performance-demanding workload, yet still provide the feature-rich environment that more traditional applications require? There are several vendors that provide this class of solution. This type of system must be designed first as a performance-focused system, then have software added to it. While the addition of that software will add some latency, it will not impact most applications. These systems typically have performance to spare. This software can be added in several ways. Some vendors provide an appliance that the performance-focused system can be connected into, allowing it to take advantage of all the features that the appliance can provide. This storage virtualization approach also allows the all-flash array to be somewhat integrated from a software services perspective. Other vendors have the ability to load storage software onto a co-processor within the flash array itself. This provides a tighter integration experience and saves the cost of an external appliance. Finally, all of these hardware-focused systems could work with any of the software-defined storage solutions that are on the market today, including those converged solutions that run within the hypervisor architecture. The PAGE 8 OF 16

key, though, is to make sure that that software-defined solution can support external, shared storage (not all do). While combining a hardware-focused solution with either an appliance or hypervisor that delivers the storage services, it s key to remember there remains one big challenge. That hardware-focused flash solution must be delivered at a price point (including software) that is in the same range as the feature-rich solutions described above. In most cases, the feature-rich solutions are still the most cost-effective, and again, 400k IOPS is more than enough for most data centers. All-flash arrays are becoming mainstream. Many vendors in the space claim price parity with performance-focused hard drive arrays. These would be arrays from name-brand vendors that are using 15K RPM drives. This claim is generally true, so any data center looking to buy a performance-focused disk array should be seriously considering an all-flash array. The choice within the all-flash segment is largely dependent on what the needs of the data center are. For most data centers, the feature-rich solutions will be all they need. But it may be worth the investigative step to confirm that and to then determine if they need a scale-up or scale-out system. PAGE 9 OF 16

FIND THE BEST SPOT FOR FLASH SSD STORAGE There are six different implementations, each aimed at reducing latency, improving IOPS and throughput, and reducing storage TCO. There are six different implementations today. Each is primarily aimed at reducing latency, improving performance in IOPS and throughput, while secondarily aimed at reducing storage total cost of ownership (TCO). This first tip will provide a brief description and reveal the pros and cons of: PCIe card(s) as cache or storage in the server. PCIe card(s) as cache in a storage system (SAN storage or NAS). HDD form factor flash SSD(s) as NAS system or storage array cache. Flash storage technology diversity requires that this subject be spread over two tips. (Here is a link to part two.) PAGE 10 OF 16

PCIE FLASH SSD STORAGE CARD(S) AS CACHE OR STORAGE IN THE SERVER. Putting the flash SSD PCIe card locally in the server on the PCIe bus puts the cache closer to the application. There is no adapter, transceiver, network cable, switch, storage controller, etc., in the path. The short distance reduces the latency, speeding up all IO operations such as reads and writes. This is why these cards are typically called application accelerators vs. storage accelerators. This type of flash SSD is primarily block. When used as cache, it requires additional software that relies upon policies to move data into and out of the cache, such as first-in, first-out (FIFO). Pros: Lowest latencies between applications and storage or storage caching. Makes a significant, noticeable and quantifiable difference for hightransactional and/or high-performance applications (OLTP, OLAP, rendering, genome processing, protein analysis, etc.) Cons: High CPU resource utilization, ranging from 5 to 25%. Relatively low capacities, (although FusionIO has a 10TB double PCIe slot card). Cards are not shareable among multiple physical servers. Each physical server requires one or more cards. Not useful for virtual servers except as cache with caching software because VM portability and resilience requires shared storage. Caching software licensing is on a per-physical-server basis. Most of the caching PAGE 11 OF 16

software is block storage, making it somewhat useless in file based storage or applications. (Nevex is the exception.) Card management is on a per-card basis, increasing administrator management tasks resulting in a high total cost of ownership (TCO). Best fits: Well-suited for high-performance compute clusters (HPC) where performance improvements in nanoseconds to microseconds are huge. Other solid fits include OLTP, OLAP, BI, social media, genome processing, protein processing, rendering, security, facial recognition, and seismic processing. PCIE FLASH SSD STORAGE CARD(S) AS CACHE IN A STORAGE SYSTEM (SAN STORAGE OR NAS). PCIe cards provide storage systems (storage vendor option) a lower cost, higher capacity, and slightly less-performing extension of the system s DRAM. It s a storage accelerator. Algorithms determine less frequently accessed data, which is quickly moved from the system s DRAM to the flash PCIe SSD cache. That cache is an extension to memory. Administrators set policies for these caches, determining what type of data should be retained or pinned in flash cache (data not evicted from the cache). Use of PCIe flash SSDs as cache reduces latency to and from the storage system by reducing disk PAGE 12 OF 16

IO when satisfying read requests and in the case of NAS, metadata as well. Pros: Reduces latencies from applications to shared storage. It works well with virtual servers, VDI, VM portability, and VM resilience. It s shareable among physical and virtual servers. It requires no server resources. Cons: Flash cache is size limited by available storage system PCIe slots. Users experience increased latencies and excessive response times because more frequent cache misses requiring requests to get the data from the HDDs. Any given storage system s flash cache cannot be shared by any other storage system. The most severe performance bottleneck is most often the storage system s CPU. As CPU utilization elevates, so does latency and user response times. Tends to be a very high or expensive TCO. Best fits: Well-suited for virtual servers and VDI. Good at providing a boost to heavy traffic applications such as Email. Does well at accelerating databases when indexes and hot files can be pinned to the cache. HDD FORM FACTOR FLASH SSD(S) AS NAS SYSTEM OR STORAGE ARRAY CACHE. HDD form factor cache is functionally similar to PCIe flash SSD storage as cache. It s a storage accelerator with similar algorithms. Instead of going into the controller as PCIe SSD cards do, HDD form factor SSDs go PAGE 13 OF 16

behind the storage controller in HDD slots. Sitting behind the controller means higher capacities but higher latencies. Pros: Reduces latency from applications to shared storage. Works well with virtual servers, VM portability, and VM resilience. It s shareable among multiple physical and virtual servers while consuming no server resources. Lower TCO per GB than the PCIe form factor. Cons: Capacities larger than PCIe flash SSDs, but limited by both flash SSD capacities and disk controller performance limitations. Users experience increased latencies and excessive response times because cache misses occur more frequently, redirecting requests to the HDDs. A storage system s flash cache cannot be shared by any other storage system. The most severe performance bottleneck is commonly the storage controller increasing latency and user response times. Best fits: Well-suited for virtual servers and VDI. Good at providing a boost to virtual and heavy traffic applications such as email. Does a good job at accelerating databases when indexes and hot files can be pinned to the cache. PAGE 14 OF 16

The next tip will provide a brief description and reveal the pros and cons of: HDD form factor flash SSD(s) as Tier 0 storage in a multi-tier NAS or storage array. HDD form factor flash SSD(s) as all SSD NAS or storage array. PCIe card(s) or HDD form factor in a caching appliance on the storage network (TCP/IP, SAN or PCIe). PAGE 15 OF 16

FREE RESOURCES FOR TECHNOLOGY PROFESSIONALS TechTarget publishes targeted technology media that address your need for information and resources for researching products, developing strategy and making cost-effective purchase decisions. Our network of technology-specific Web sites gives you access to industry experts, independent content and analysis and the Web s largest library of vendor-provided white papers, webcasts, podcasts, videos, virtual trade shows, research reports and more drawing on the rich R&D resources of technology providers to address market trends, challenges and solutions. Our live events and virtual seminars give you access to vendor neutral, expert commentary and advice on the issues and challenges you face daily. Our social community IT Knowledge Exchange allows you to share real world information in real time with peers and experts. WHAT MAKES TECHTARGET UNIQUE? TechTarget is squarely focused on the enterprise IT space. Our team of editors and network of industry experts provide the richest, most relevant content to IT professionals and management. We leverage the immediacy of the Web, the networking and face-to-face opportunities of events and virtual events, and the ability to interact with peers all to create compelling and actionable information for enterprise IT professionals across all industries and markets. PAGE 16 OF 16