The next version of NetApp® ONTAP® is right around the corner, but that doesn’t mean we’re done with ONTAP 9.2 yet. In fact, we’re setting records with it! For certain workloads you can get significantly more out of your storage systems by deploying ONTAP 9.2 (or newer versions) and using the FlexGroup feature for scalable, high performance data containers.
NetApp recently submitted SPEC SFS®2014_swbuild test results using NetApp FlexGroup volumes and achieved an enterprise-leading 1.04 overall response time (ORT) score. The SPEC SFS®2014_swbuild test is a benchmark for software build workloads, which characteristically include many files, high ingest rates, and a large amount of metadata operations. Not coincidentally, NetApp FlexGroup volumes were designed for that type of workload, and it showed in the benchmark results.
The testing was done with a normal hybrid FAS8200 cluster with SAS drives. The test setup included:
- ONTAP 9.2
- Four-node FAS8200 hybrid flash cluster (not All Flash FAS)
- 72 4TB 7200 RPM 12Gb SAS drives per HA pair
- 20 IBM servers/clients
- 10GbE network (four connections per HA pair)
The NetApp ORT benchmark result underscores the fact that NetApp is the best choice for customers with demanding engineering and design workloads. As organizations move to a DevOps practice for continuous integration, the number and frequency of software tests and builds are increasing rapidly, requiring a scalable, high-performance data infrastructure.
Figure 1 shows the ORT comparison to other storage system submissions. ORT measures the storage system’s ability to react to workloads as fast as possible.
Along with the 1.04 ORT, FlexGroup volumes also delivered the best throughput of the submissions, with over 4000MBps in a four-node FAS8200 cluster, as shown in Figure 2.
Throughput metrics show how much data can be transmitted over the wire and processed by a storage system. Throughput is a good way to measure the performance for streaming audio or video workloads.
Although throughput is a nice stat, many applications (and their owners) care more about latency. Latency is the measure of time it takes for a storage system to respond to a client or application on a read or write request. Fortunately FlexGroup volumes had that covered, too, as illustrated by Figure 3.
NetApp FlexGroup volumes were also able to handle a larger scale of builds, reaching 520 builds, while only incurring a maximum latency of 3ms. Number of builds measures the number of simultaneous jobs the storage system can handle before it starts to max out. No other submission made it past 500 builds, and even then, it was at a more severe latency cost. (See Figure 4.)
These combined results show that with ONTAP and NetApp FlexGroup volumes, you don’t have to decide between IOPS, latency, throughput, or the number of builds a cluster can handle. You get it all. In addition to performance, you also get:
- Cost benefits of storage efficiencies
- Performance boosts without necessarily needing to go all flash
- Flexibility to go all flash if you need it
- Nondisruptive scale-out and scale-up of file systems and clusters to grow with your applications
- Enterprise-class Snapshot™ technology
- Simple, streamlined deployment
For more information, see the following references:
- TR-4557: FlexGroup Technical Overview
- TR-4571: FlexGroup Best Practice Guide
- TR-4617: EDA Best Practices
- Tech ONTAP Podcast Episode 46: FlexGroup Volumes
- FlexGroup Volumes: An Evolution of NAS
- If you’ll be at NetApp Insight 2017, come find me in session 16594-2
You can also e-mail us at firstname.lastname@example.org.