Rescue implementation differs wildly between hypervisors
Bug #829020 reported by
Soren Hansen
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Fix Released
|
Wishlist
|
Vish Ishaya |
Bug Description
The implementation of the rescue call differs wildly between at least libvirt and XenAPI (I didn't check the others). XenAPI uses a pristine copy of the original image from which the instance was created, while the libvirt driver uses a specific rescue image (defined by a flag on the compute node) regardless of which image the instance is based off of.
We need to fix this. There's no reason these drivers should be behaving so differently.
Related branches
lp:~vishvananda/nova/lp829020
- Dan Prince (community): Approve
- Devin Carlen (community): Approve
-
Diff: 132 lines (+40/-37)1 file modifiednova/virt/libvirt/connection.py (+40/-37)
lp:~vishvananda/nova/lp829020-rbp
- OpenStack release team: Pending requested
-
Diff: 132 lines (+40/-37)1 file modifiednova/virt/libvirt/connection.py (+40/-37)
Changed in nova: | |
importance: | Undecided → Wishlist |
status: | New → Confirmed |
tags: | added: consistency |
Changed in nova: | |
assignee: | nobody → Vish Ishaya (vishvananda) |
milestone: | none → 2011.3 |
Changed in nova: | |
status: | Confirmed → In Progress |
Changed in nova: | |
status: | In Progress → Fix Committed |
Changed in nova: | |
status: | Fix Committed → Fix Released |
To post a comment you must log in.
Would a single rescue image libvirt style work if you're running multiple OSes? Is a linux rescue image appropriate for rescuing a windows image?