I have heard and read some “interesting” comments regarding the recent announcement by 5 SNIA members companies (EMC, HP, Sun, Symantec and Hitachi Data Systems) about the SMI-S Advanced Services. Many of these comments were highly speculative and not often reflecting the approach the 5 vendors decided to take. This announcement got nick-named “Anti-Aperi” or even “Anti-IBM”…but again that’s not the case and missing the point of this announcement.
So what is this all about? Fairly simple: Managing storage resources remains a challenging exercise for many IT departments, especially for those working in heterogeneous environments; i.e. most of them. Standards like SNIA SMI-S do help to provide management tools with multi-vendors management capabilities; however its development & adoption cycles by vendors and users are longer than initially anticipated, creating somehow a need to accelerate the next development steps for SMI-S, leveraging the significant efforts from SNIA and the Storage Management Forum.
SNIA has provided a good framework for SMI-S to become a device-level interface with the right abstraction layer for modelling, discovering and controlling multi-vendor storage network equipment. However storage users have also told the storage vendors that management functions need to be more homogeneous across the board, possibly beyond storage, with more common management features/functions. I.e. SMI-S need to provide reference building blocks for advanced management features, transitioning SMI-S in to a system level interface. Benefits for vendors, ISV and users will include the quicker delivery of management services for heterogeneous resources, closer to business application requirements. Clearly this announcement is here to boost the discussion around the future of SMI-S.
So what about coding or code contribution(s)? Well before any coding you need strong specifications, especially if you seriously think about providing reference elements. SNIA can look in to the distribution model for a reference code at a later stage. Right now, besides organizational talks, the focus of this effort will be on the specification objectives and content. No rush to coding…
So if this announcement shares some goals with what Aperi had announced, the intended path and framework (SNIA) for this effort is indeed different. All SNIA members are invited to join us!
