[COE-43] veth interface error: "could not open network device veth9e3d3ff1 (No such device)" Created: 05/Dec/18 Updated: 23/Jan/19 Resolved: 21/Jan/19 |
|
| Status: | Verified |
| Project: | coe |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Medium |
| Reporter: | Michael Vorburger | Assignee: | Faseela K |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Description |
|
In
|
| Comments |
| Comment by Michael Vorburger [ 05/Dec/18 ] |
|
sudo ovs-vsctl del-port veth741cafb2 let's one manually delete such ports. We suspect that they are for some reason left over after a kubectl delete pod (see also COE-45, although k.faseela says that "that exception will not cause port not to be deleted from switch"). |
| Comment by Michael Vorburger [ 13/Dec/18 ] |
|
serngawy, on IRC k.faseela suggested that I assign this to you... |
| Comment by Michael Vorburger [ 13/Dec/18 ] |
|
COE-45 may be related to this? |
| Comment by Mohamed ElSerngawy [ 18/Dec/18 ] |
|
This happen because of changing the iface-id to (cluster-id and podname) instead of (namespace and podname) check [0]. I don't really understand what is the purpose of doing that, you still can create 2 pods with the same name in the same cluster under 2 different namespace. "cluster-id and podname" is not a unique identifier. [0] https://github.com/opendaylight/coe/blob/master/odlCNIPlugin/odlovs-cni/ovs_odl.go#L196 |
| Comment by Faseela K [ 21/Jan/19 ] |
|
Karthika_rp : Could you please verify if the issue is not reproducible on latest master CSIT and close this JIRA? |