InfoWorld review: Dell iSCSI SAN sizzles with SSD, dynamic storage tiering

The Dell EqualLogic PS6010XVS with automated storage tiering leverages fast SAS or faster SSD storage depending on workload

By Paul Venezia, InfoWorld |  Storage, Dell, SAN

Dell EqualLogic PS-series SANs have long led the iSCSI pack with excellent performance, good management tools, 10G Ethernet support, and the ability to scale capacity and I/O simultaneously by stacking together appliances. The latest addition to the line offers more reasons to take a close look, with a high-performance solid-state drive (SSD) array, dynamic storage tiering, and tighter integration with VMware.

The new Dell PS6010XVS model differs from its 16-drive, 4U PS-series brethren only on the inside, where it houses eight 100GB Samsung SSDs and eight 450GB 15K RPM SAS drives, each run in their own RAID6 array. By itself, the SSD-SAS combination isn't terribly interesting -- but what the new version of the EqualLogic firmware can do with it is. 

[ Use server virtualization to duplicate resources and create failover capabilities at a cost far below that of conventional solutions -- without compromising reliability. Download InfoWorld's High-Availability Virtualization Deep Dive Report. ]

Version 5 of the EqualLogic firmware brings a host of new features to the EqualLogic storage line, but the major benefit to the Dell PS6010XVS is the dynamic storage tiering capability. In traditional tiered storage configurations, volumes are created on disparate storage technologies, SAS, SATA, and perhaps SSD, and data is placed on those volumes according to how heavily that info will be accessed. For example, databases might reside on the SAS or SSD arrays, while the cheaper SATA arrays house data that doesn't get much attention. Basically, admins make an educated guess as to the usage patterns of the data in their care and place it accordingly. Most times they're on the right track, but it's impossible to be sure that you have everything tuned correctly.

EqualLogic's new code keeps tabs on how data is being accessed at the block level and shifts heavily accessed portions to the SSD storage on the fly, increasing I/O substantially in some cases. As the workload shifts, it will return some blocks to SAS storage, other blocks to the SSD, and so forth. There's no more guesswork involved; it all happens automatically based on actual production workloads.


Originally published on InfoWorld |  Click here to read the original story.
Join us:
Facebook

Twitter

Pinterest

Tumblr

LinkedIn

Google+

Answers - Powered by ITworld

Ask a Question