Unable to get device_path from network mounted volume in extend in use
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Invalid
|
Wishlist
|
Jose Castro Leon |
Bug Description
While implementing support for extending in-use volumes for RBD, I've hit a bug as the code relies on the server having access to the devices that are visible in its filesystem through the device path. On network attached volumes in which the qemu process directly access the device over the network, this device_path attribute does not exist that leads to an exception.
https:/
...
state = guest.get_
if active_state:
dev = guest.get_
...
Changed in nova: | |
importance: | Undecided → Wishlist |
tags: | added: ceph libvirt |
Fix proposed to branch: master /review. openstack. org/590188
Review: https:/