Analysis

posted on 03 March 2008 08:40

Andrew Monshaw, IBM's GM for system storage, has said the IBM file storage future is an XIV one of clustered inexpensive storage boxes.

In an InformationWeek interview Monshaw implied that the DSx000 range was facing a next-generation architecture based on the XIV acquisition. No time-scale was given and no mention whatsoever was made of DSx000 replacement. If IBM successfully introduces an XIV-based file storage line than file storage applications could be expected to move from DSx000s to the new line.

The background is as everybody suspected; rampant unstructured file information growth and the need for greatly expanded scalability in storage system performance, capacity and the ability to protect data.

The DS range will get a solid-state disk (SSD) infusion with, as with EMC's Symmetrix DMX4, SSD being used to provide a new top tier of performance. Monshaw said that SSD's limited write cycle life was an issue and IBM was developing algorithms to extend it to a three to five year period. The term 'wear-levelling' was not used but it is probably what is being developed.

We might reasonably expect an announcement later this year which could refer to both the DS8000 and the DS6000.

Monshaw also gave a picture of the external disk storage market which he said was worth $4.4 billion and was led by EMC (22 percent share), IBM (14 percent), HP (13.4 percent), Dell (9.3 percent), and HDS (8.5 percent). The HDS share is actually higher as HP's top-end EVA arrays are sourced from HDS. Perhaps HDS has a true 10-11 percent share.

In terms of where the NetApp-sourced N series sits in IBM's storage product plans one interpretation of what Monshaw said is that NetApp is, crudely-speaking, a stop-gap supplier.

Another could be that the DS8000 and DS6000 range will be focussed on mainframe-attach storage. The DS4000 and DS3000 will be gradually replaced by NetApp's N series with the XIV technology being used for a new highly-scalable file storage product line where NetApp is is not a presumed viable supplier in this interpretation of IBM's view.