DRBD transport layer to DRBD Cinder nodes
We propose to implement a block storage driver for Nova that uses the DRBD network protocol directly to access DRBD 9 storage servers, instead of needing an iSCSI hop.
Blueprint information
- Status:
- Complete
- Approver:
- Matt Riedemann
- Priority:
- Low
- Drafter:
- Ph. Marek
- Direction:
- Approved
- Assignee:
- Ph. Marek
- Definition:
- Approved
- Series goal:
- Accepted for pike
- Implementation:
-
Implemented
- Milestone target:
-
pike-2
- Started by
- Matt Riedemann
- Completed by
- Matt Riedemann
Related branches
Related bugs
Sprints
Whiteboard
Please submit a nova-spec for this that references this blueprint. Many thanks. --johnthetubaguy 19th November 2014
You should not set a milestone target unless the blueprint has been properly prioritized by the project drivers.
(This is an automated message)
Addressed by: https:/
This is just adding a brick connector to the libvirt driver in Nova, so we don't need a spec for this. The Cinder and os-brick changes landed in Mitaka, the Nova change just needs work. Also, it can't add a new compute rootwrap filter, it has to use oslo.privsep. Angus Lees could help with that if the spec doesn't make it clear on how to do the privsep changes. There was also an etherpad on privsep from the Austin summit: https:/
-- mriedem 20160519
Gerrit topic: https:/
Addressed by: https:/
Connecting Nova to DRBD storage nodes directly
There are still pending changes for this and we're now at non-priority blueprint feature freeze for Newton. -- mriedem 20160701
I've re-approved this for Pike as there is interest in reviving it. The cinder and os-brick changes merged in Mitaka. flip214 is actively monitoring the DRBD CI for Cinder. -- mriedem 20170503
Work Items
Dependency tree
* Blueprints in grey have been implemented.
