Big Blue Rolls-Out Features For Systems Storage

During the weeks of April 12th and 19th, IBM announced the release of new features and functions for the IBM System Storage DS8700 and XIV, along with other enhancements and new products in their overall storage line up. Details of the whole announcement and a video with Doug Balog, IBM Storage Platform VP and Disk Storage Business Line Executive, can be located by clicking here.

Tom Trainer

May 3, 2010

6 Min Read
Network Computing logo

During the weeks of April 12th and 19th, IBM announced the release of new features and functions for the IBM System Storage DS8700 and XIV, along with other enhancements and new products in their overall storage line up.  Details of the whole announcement and a video with Doug Balog, IBM Storage Platform VP and Disk Storage Business Line Executive, can be located by clicking here.

In my opinion, from a disk-storage perspective, the most exciting piece of the April storage announcements is the release of IBM Easy Tier for the DS8700.  Easy Tier is a sub-LUN/volume data management and movement methodology that competes head-to-head with offerings from Compellent, 3Par and EMC, who has made public statements that their sub-LUN offering (FAST2) will be available in the Fall of this year.

So, you may ask, "What's the big deal with sub-LUN data management and movement capability?  And you may say, "Hey, you wrote about this a few weeks ago when 3Par released their Adaptive Optimization functionality."  Well, yes, I did write about 3Par and their release and that I felt it as being the first-to-market with sub-LUN data management in the enterprise storage space. With IBM's release of Easy Tier, they take the second place spot and third overall, behind Compellent and 3Par with sub-LUN data management capabilities now built in to the operational software of arrays. I anticipate both IBM and 3Par having their respective sub-LUN offerings operating within production data centers before the beginning of June, 2010.

sublun.pngThe big deal with sub-LUN data optimization is that smaller, more discrete quantities of data - rather than a whole volume or LUN - can be monitored, identified and moved based on the access characteristics of the specific data quantities. Each vendor has a name for their functionality and a naming convention for the data quantities that it monitors and then automatically moves to higher performing (or lower performing) storage.  The recently announced IBM  Easy Tier identifies their quantities of data under access measurement as "Extents."  

The even bigger deal here is the business impact of real-time data management.  Once data is free to be accessed in a faster manner by a host/applications, say from a Solid State Disk (SSD), the host/applications can run faster, and if the application is a revenue-generating one, then it stands to reason that more revenue can be generated in a period of time with faster disk I/O operations. Lets not forget that while SSDs may still be a pricey disk solution, their use in combination with spinning disk can dramatically improve the overall energy consumption characteristics of a storage array, saving you money on electricity usage and air conditioning.IBMsubVolDiagram.pngConversely, as access to data sitting on SSDs begins to wind down, and the frequency of access declines, these new data-management solutions will move the data back to slower performance spinning disks such as Fibre Channel or SATA drives. It's not to much of a stretch to anticipate storage vendors or smart back-up and archive software vendors jumping on the sub-LUN data-management bandwagon and developing software that leverages API's to have the storage array manage data from highly accessed SSDs to slower Fibre Channel or SATA disk drives and then right on out of the storage array to an archival platform - or out to the Cloud for re-purposing or long-term retention.

As mentioned previously, the IBM Easy Tier solutions manages data Extents on volumes.  These monitored extents form a virtualized Extent environment, and the extents are RAID-protected and are moved either automatically or manually between SSD and spinning disks based on frequency of access.  Figure 1 is a simple illustration of the IBM technique.

We at Analytico have started tracking storage vendors as they make public comments about their sub-LUN data management capabilities, or future offerings. From our perspective, it is easier to track the vendors and their offerings in table format. The vendors we are currently tracking include Compellent, 3Par, IBM, EMC, Hewlett Packard (HP) NetApp, and Hitachi Data Systems (HDS). We provide information in Table 1 as a sample containing the kinds of specific vendor information we are tracking. Please feel free to comment on this blog and suggest additional areas you would like to see tracked.
 
Other Related Disk Storage Announcements by IBM
IBM Storage Tier Advisory Tool Monitor and Measurement offer analytics for use with IBM Easy Tier on the DS8700. IBM XIV Storage System adds support for higher density, 2TB, SATA disks for near-line storage capacity increases, and some times offline via "spin down" CLI command support. This is a natural storage capacity addition for the XIV. It also includes new, low-voltage, components in its green initiatives discussions and improving the power consumption of the box is a step in the right direction. Remote mirror of snapshots provide users with the ability to execute a snapshot copy at the remote site and is a big step forward in supporting consistent splits in support of consistency groups in high availability XIV installations.

All of the major storage vendors will include FCoE support within their arrays throughout 2010 and early 2011.  If uncertain, users should ask their storage vendors what the FCoE support time-line looks like. Overall, most agree that few will deploy FCoE connectivity in 2010.  However, there will be greater adoption in 2011 as sever vendors begin to ship Converged Network Adapters (CNAs) and LAN On Motherboard (LOM).  Also, the industry believes that full-scale adoption and deployment will be in full swing during 2012.  IBM is correct to deploy FCoE on XIV now in preparation for future user demands.

I was fortunate to be able to attend the IBM Storage Symposium in Boston in October 2009.  At that symposium IBM was clear in their intent to keep a positive momentum going with the IBM System Storage products. IBM vowed to continue to innovate and to provide sub-LUN level data management and movement within its products. With its strong background in data set (sub volume) management in the mainframe space, it seemed only a natural that IBM would be able to build a sub-LUN data management capability for its open connectivity side of the business on the DS8700. Perhaps more challenging may be the implementation of sub-LUN data-management on the XIV and SVC. I believe these capabilities are possible on these products. However, much like EMC, the data-management techniques may vary greatly based on the differing architectures spanning the product lines.

In my opinion, the bottom line here is that IBM does not intend to be a shrinking violet in the storage industry, and these announcements continue to provide evidence that IBM will play catch up if it has to, but fully intends to keep moving forward with the development of innovative storage solutions. Now, IBM, it's time to innovate even smarter and start to take those leaps and bounds forward versus the competition.

If you would like to see more areas of sub-LUN characteristics V tracked or other feature functions tracked across vendors, post a comment to this blog and let me know your desire.  Also, I will be releasing a more detailed report on sub-LUN data optimization in May of 2010.  I will be sure to comment on its availability when published.  At the time this blog was written Tom Trainer had no business relationship planned, or in process, with IBM.
 

Read more about:

2010

About the Author(s)

SUBSCRIBE TO OUR NEWSLETTER
Stay informed! Sign up to get expert advice and insight delivered direct to your inbox
More Insights