Randy Bias wrote:
Just want to clarify re: storage that we don't need it in the initial draft, but are keen to have a standard around some aspects of it. If I recall correctly we drew the line on storage saying that the core API functionality would allow managing it at the infrastructure (disk, mount, etc.) level, but not necessarily at the file/object level. We'll also need the latter, which is what I believe the SNIA folks are working on. I can't recall the particular working group. Mark Carlson, do you know which one it is?
The SNIA Cloud Storage TWG is the central organizing point for all the APIs. There are other TWGs with related work such as SMI-S, Long Term Archive, XAM and Hypervisor Storage APIs. There is a Google group that is available to non-members as well: http://groups.google.com/group/snia-cloud/subscribe
Anyway, in order of priority my desired list of deliverables looks something like this:
1) Clean core API + extensibility delivered ASAP (preferably by end of June) 2) Storage mgmt extensions for block devices and mount points (end of July) 3) File/object based extensions to allowing send/receive and arbitrary URLs for retrieval of any given file/object (end of August)
Just putting my desired scenario and objectives here. I realize we might not be able to achieve all of these in this time frame. I also realize that SNIA probably is the best owner for 3. I'm eager for them to join us on the next call.
It'll be the May 13th call. -- mark