Sun storage thoughts
Quarterly announcements, roadmaps and utility storage
By Chris Mellor | Published: 10:30, 05 November 2005
Storage customers better get used to Sun's quarterly storage announcements. The next one is due in December and, in Sun terms, the Titanic announcement was out of schedule, something to add to the Forum meeting's significance, and it was a very significant meeting, one where StorageTek's tenants, as it were, met their new landlord.
Mark Canepa, Sun DMG head, and Nigel Dessau, Sun DMG's tape head, emphasised Sun's commitment to the StorageTek mainframe market. They also revealed some details of Sun's storage roadmap. More information has also come available about Sun's utility storage activities.
The story here is that StorageTek tapes play the kingpin role while Sun's drive arrays rule the primary storage roost:-
- The StreamLine 8500 is and remains the flagship tape library. It will be developed and enhanced
- The Sun C10 mid-range tape library, will be end-of-lifed.
- The Sun C2 and C4 drives will remain as they slot in below StorageTek's automated tape products.
- The FlexLine 680, intended by StorageTek as a frontline array will be re-positioned as nearline storage.
- The Flexline 380 disk array overlaps with Sun's 6xxx StorEdge products and will most likely be end-of-lifed.
- The FlexLine 240, and StorEdge 6130 will be likely replaced by a single product in 2006.
- The mainframe VSM product will remain as a flagship product and be enhanced.
- The long-awaited open systems virtual tape product weill be accelerated and should appear later this year.
- StorageTek's GSM storage management product will be integrated into Sun's StorEdge Enterprise Storage Manager.
Where other vendors are OEM'ming ISV virtual tape products, such as IBM with FalconStor, StorageTek has been developing its own, based on its mainframe virtual tape product, and this make rather than buy exercise has proved costly, particularly in terms of time.
Utility storage - one-stop data parking
What they probably didn't say much about at all was Sun's utility storage idea. Our companions at The Register write that this, along with the utility computing idea, has been pretty much a dead loss apart from much learning by Sun, presumably with the main lesson being that it hasn't worked at all.
There isn't a single signed up customer for its $1/GB utility storage concept. However Sun's utility computing head, Aisling MacRunnels, is reported as saying she has thousands of customers lined up and ready to go for the utility computing idea, but she said nothing about the utility storage idea.
It occurs to me that businesses don't just want a place to park data. They want to move data between various parking places - frontline, nearline and archive storage for example, and need storage management software. The Sun storage offer is basically a one-stop data parking exercise. It's far too simplistic an idea for a storage utility which should, I think, offer a complete storage utility, if it is possible to craft a business offer around something that doesn't exist yet.
Obviously it isn't, so I expect Sun's utility storage offer to be substantially altered. We'll probably here more in Sun's next quarterly announcement-fest in December.
The dog that didn't bark
We didn't get much information on the direction of Sun's storage virtualisation and management strategy. For example, the Enterprise Storage Manager (ESM) is the storage management product. It has to deal with two multi-vendor storage array virtualisation products: the Hitachi Tagmastore and NCS55; and the 6920. Where do the StorageTek arrays fit here? Where will iSCSI arrays fit? Will iSCSI array virtualisation be offered? How will it be managed? How will ESM be integrated with StorageTek's ILM concepts? How will NAS fit in here? Where does Honeycomb fit?
There is much, much work still to be done to evolve a single unified storage strategy from Sun and StorageTek's product plans and individual strategies. With the product line-up and engineering resource the combined company has there is the prospect of a compelling vision and strategy emerging.
Let's hope it does and that EMC, HP, IBM and NetApp get a really powerful competitor.