I read with interest yesterday’s announcement from IBM that they are to acquire Storwize, a vendor of NAS compression technology. Maybe I don’t understand enough about the technology, however I don’t see much benefit in installing an application in front of my NAS environment to only achieve between 50-90% reduction in storage. What’s more surprising is that IBM would want to acquire this technology.
Firstly, let’s summarise the benefits of the Storwize offering:
- Better Storage Utilisation
- Lowers Capital and Operational Costs
- Better Energy Efficiency
These bullet points are taken directly from the Storwize website and are then expanded into more detail. Whilst implementing compression may save some storage space in the short term, it doesn’t address the overall reasons for growth within an organisation and at some stage capacity will reach previous levels, regardless of the ability to compress that data. I’d also agree that compression reduces some capital costs – but these will be via cost avoidance rather than the ability to remove existing hardware. This implies that the savings can only be made if significant growth exists in the environment in the first place.
Implementing compression and de-duplication is analagous to the person who won’t face up to their debts. Each month they find other ways of avoiding bankruptcy; take out a new loan, extend credit, use one credit card to pay off another. Eventually the house of cards comes falling down. This may be a dramatic comparison but in the storage world, with data compression, the growth rate still continues; the underlying problem hasn’t been addressed.
So what about the negatives? Well, firstly, all data needs to go in and out of the appliance. This immediately puts in place restrictions and bottlenecks. It also creates another layer of compatibility and support. It also introduces scalability issues that then needs to be addressed by implementing multiple devices; all of a sudden a different management problem arises. Here’s another thought; if the appliance is removed, does all data need to be “unpacked’? If the underlying NAS environment is replaced, does the data have to come in and out of the appliance as it is moved to the new location?
So, now to what seems more surprising; acquisition by IBM. From my experience, IBM has plenty of knowledge and expertise in data compression technologies, stretching back decades; think of IDRC in 3480 tape cartridges as a simple example. Why then do IBM feel the need to acquire this technology? In addition, where will IBM sell this technology – in front of Netapp rebranded filers? How does that square with the use of ASIS de-duplication? Is this acquisition implying ASIS isn’t fit for purpose?
Only time will tell if this is a sound acquisition of the demise of another startup; I’m sure the guys at Storwize are pleased; they have had their idea validated and no doubt made some serious money. Kudos to them…