Status: Accepted
Owner: [email protected]
Labels: Type-Defect Priority-High Milestone-Release2.8
New issue 680 by [email protected]: Cancelling jobs fails
http://code.google.com/p/ganeti/issues/detail?id=680
What steps will reproduce the problem?
1. Ready an empty cluster for instance creation; have at least two nodes.
2. Through the command line, issue two longer-lasting gnt-instance add
commands while using the --submit flag, e.g.
gnt-instance add -o debian-image -t drbd -n node1:node2 -s 5g --submit
instance1
-> JobId: jobid1
gnt-instance add -o debian-image -t drbd -n node1:node2 -s 5g --submit
instance2
-> JobId: jobid2
3. Cancel the second job
gnt-job cancel jobid2
What is the expected output? What do you see instead?
As the second job is queued behind the first one and waiting, it should be
possible to cancel it. It does show up as being in the process of being
cancelled, yet the cancellation never happens, and the job runs
unexpectedly.
Snippets from two invocations of gnt-job list:
...
jobid1 running INSTANCE_CREATE(instance1)
jobid2 canceling INSTANCE_CREATE(instance2)
...
...
jobid1 success INSTANCE_CREATE(instance1)
jobid2 running INSTANCE_CREATE(instance2)
...
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
