Fibre Channel Multipath attach race condition
Bug #1175366 reported by
Walt Boring
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Fix Released
|
Medium
|
Walt Boring |
Bug Description
When the system is under load and a Fibre Channel attach happens with multipath installed, sometimes only 1 devices shows up under multipath -l command. This causes the detach to fail to detach all of the devices since the libvirt volume code runs multipath -l at attach time, and not all devices show up at that time due to system load and/or FC fabric traffic.
We need to rediscover all of the devices at detach time and ensure we are removing all of those devices. The devices should exist at detach time.
Changed in nova: | |
status: | New → Triaged |
importance: | Undecided → Medium |
Changed in nova: | |
status: | Triaged → In Progress |
Changed in nova: | |
milestone: | none → havana-2 |
status: | Fix Committed → Fix Released |
Changed in nova: | |
milestone: | havana-2 → 2013.2 |
To post a comment you must log in.
Here are example output for multipath -l /dev/sdl at different times.
ATTACH time
350002ac110e2383d dm-5 3PARdata,VV
size=4.8G features='0' hwhandler='0' wp=rw
`-+- policy='round-robin 0' prio=-1 status=active
`- 2:0:0:5 sdl 8:176 active undef running
DETACH TIME
350002ac110e2383d dm-5 3PARdata,VV
size=4.8G features='0' hwhandler='0' wp=rw
`-+- policy='round-robin 0' prio=-1 status=active
|- 2:0:0:5 sdl 8:176 active undef running
`- 1:0:0:5 sdm 8:192 active undef running