Briefing Note: NetApp upgrades EF Arrays to address Performance First Workloads

Flash storage is being used in three areas of the data center infrastructure. First, flash and all-flash arrays are seeing adoption in both server and desktop computing environments, as well as mid-range database applications. Second, for databases that demand extremely high performance storage, a different type of flash system is used, a flash appliance. This is a shared flash system that focuses less on delivering features and more on delivering performance. Third, there is the use of flash storage in converged and hyper-converged architectures. As we explained in our article “What are Converged Infrastructures?” flash can be deployed again in a shared all-flash array or internally inside the compute servers.

Addressing the Performance First Category

The first category of flash storage use typically has hundreds if not thousands of workloads all demanding high performance (50K IOPS+). Flash helps because it can provide performance to the aggregate of all the workloads. NetApp addresses this section of the market with its FAS and AFF storage systems. The second category, “performance first”, is typically a single application that needs extremely high performance (500k+ IOPS). NetApp addresses this category with its EF/F series of arrays.

New NetApp E560 and E5600

In its recent announcement, NetApp upgraded its capabilities to better address the second flash category with the new EF560 all-flash array and the E5600 hybrid array. It is fair for NetApp to refer to these as “arrays”, instead of “appliances”, because while not as feature laden as NetApp’s FAS systems they do deliver plenty of enterprise class storage services.

Extreme Performance Needs Scale Up Architectures

It is important not to be fooled by scale-out storage systems that claim millions of IOPS. Typically these systems justify their claims by aggregating performance across many workloads. They would get to 500k IOPS by combining the performance delivered to ten 50K workloads, which is fine for an environment with many parallel workloads. The problem is that many of the “performance first” category workloads are single applications and even single volume in nature. They often need one very fast scale up storage system.

Better Storage System Design Leads to Better Performance

The common denominator in shared all-flash and flash appliances is the media. While the actual performance of the NAND flash devices themselves can vary from supplier to supplier, the gating factor is the speed at which the appliance can ingest data and move that data to and from these flash storage devices. In this announcement from NetApp these new systems increase IOPS performance by 25%. The EF560 and E5600 are based on new platforms with faster, multicore CPUs.

More importantly, SANtricity 8.2, the EF storage operating system, has been upgraded to provide multi-threading, essential for getting maximum performance out of a multi-core processor. Most vendors without multi-core support are forced to dedicate each core to a specific function, which leads to an imbalanced use of the cores.

In addition to a faster platform and an upgraded storage operating system, NetApp also increased performance. The systems now provide new connectivity options with 8 x 12Gb SAS connections and 4 X 56Gb InfiniBand connections. This is in addition to upgrading standard connections to either 8 x 16Gb FC or 8 X 10GbE iSCSI.

The Results

All of this work leads to an E560 system that NetApp claims can deliver a sustained 650k IOPS at less than 800 Microseconds of latency. While there are faster systems on the market, the E560 is near the top end, from a latency perspective, which for many environments is more relevant. A proof point is that the E560 had an excellent showing in a recent SPC benchmark.

Those few systems that are faster (in IOPS) do so with essentially no features. SANtricity provides snapshots, mirroring, replication and encryption. It also provides application plug-ins for Oracle, MS-SQL, VMware and Splunk.

NetApp also makes a point that these systems do NOT have deduplication and/or compression. They feel that customers in the extreme performance category are less concerned about space efficiency, and more concerned about performance and capabilities.

StorageSwiss Take

There are many data centers where this level of extreme performance is simply not needed. For those environments NetApp has their FAS series of products. But unlike other vendors NetApp can also satisfy customers who are at that extreme end of the performance spectrum. Few vendors can participate in both of these categories and do so well.

Click Here To Sign Up For Our Newsletter

George Crump is the Chief Marketing Officer at VergeIO, the leader in Ultraconverged Infrastructure. Prior to VergeIO he was Chief Product Strategist at StorONE. Before assuming roles with innovative technology vendors, George spent almost 14 years as the founder and lead analyst at Storage Switzerland. In his spare time, he continues to write blogs on Storage Switzerland to educate IT professionals on all aspects of data center storage. He is the primary contributor to Storage Switzerland and is a heavily sought-after public speaker. With over 30 years of experience designing storage solutions for data centers across the US, he has seen the birth of such technologies as RAID, NAS, SAN, Virtualization, Cloud, and Enterprise Flash. Before founding Storage Switzerland, he was CTO at one of the nation's largest storage integrators, where he was in charge of technology testing, integration, and product selection.

Tagged with: , , , , , , , , , , , , , , ,
Posted in Briefing Note
One comment on “Briefing Note: NetApp upgrades EF Arrays to address Performance First Workloads
  1. […] Briefing Note: NetApp upgrades EF Arrays to address Performance First Workloads  by George Crump at […]

Comments are closed.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 25,553 other subscribers
Blog Stats
%d bloggers like this: