cloudstack-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mike Tutkowski (JIRA)" <j...@apache.org>
Subject [jira] [Created] (CLOUDSTACK-8341) GSoC 2015
Date Mon, 23 Mar 2015 17:15:11 GMT
Mike Tutkowski created CLOUDSTACK-8341:
------------------------------------------

             Summary: GSoC 2015
                 Key: CLOUDSTACK-8341
                 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-8341
             Project: CloudStack
          Issue Type: Improvement
      Security Level: Public (Anyone can view this level - this is the default.)
          Components: Management Server, VMware
    Affects Versions: Future, 4.6.0
         Environment: All OSes the CloudStack Management Server is expected to run on.
            Reporter: Mike Tutkowski
             Fix For: Future, 4.6.0


ACS users could benefit from the ability to take snapshots (for managed storage) that reside
on primary storage (instead of backups that reside on secondary storage).

Benefits could include the following:

* Fast snapshots instead of slower backups
* Make use of compression, de-duplication, and thin provisioning that may be supported on
the primary storage

Some work has already been done with regards to this topic using the XenServer hypervisor.
However, that hypervisor currently has restrictions around immutable UUIDs on its SRs and
VDIs that effectively limit the value proposition of backend snapshots.

Let's target VMware for this feature as it has the ability to resignature its datastores and
thus make effective use of backend snapshots of datastores.

Related: https://pubs.vmware.com/vsphere-50/index.jsp?topic=%2Fcom.vmware.wssdk.pg.doc_50%2FPG_Ch8_Storage.10.9.html

This work may entail modifying components in the management server that expect snapshots to
only reside on secondary storage.

In addition, the VMware agent code will need to be updated.

Also, a custom snapshot strategy and a custom data motion strategy will either need to be
created (or the code will have to be placed inside of an existing snapshot strategy and data
motion strategy).

Once the snapshots can be created on primary storage, we need to make sure we can create volumes
and templates from them.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message