Bug #3775
deleted VMs are canceled asynchronously, host/ip resources already shown as freed before releasing them
Status: | Closed | Start date: | 04/27/2015 | |
---|---|---|---|---|
Priority: | Normal | Due date: | ||
Assignee: | - | % Done: | 0% | |
Category: | Core & System | |||
Target version: | Release 5.0 | |||
Resolution: | fixed | Pull request: | ||
Affected Versions: | OpenNebula 4.12 |
Description
When deleting big number of VMs in a host we found, that VMs are canceled asynchronously.
We saw host resources and/or IP adresses freed when VM is still running.
Users could submit new VMs using those resources still in progress to be freed.
Related issues
Associated revisions
History
#1 Updated by Carlos Martín about 6 years ago
- Related to Feature #3654: Implement extra VM state transitions added
#2 Updated by Ruben S. Montero about 6 years ago
- Target version set to Release 4.14
#3 Updated by Ruben S. Montero about 6 years ago
- Status changed from Pending to Closed
- Resolution set to worksforme
Hi Rolandas,
If you need to sync the delete of a VM use the shutdown method. Delete is mend to be asynchronous and to be used in very special situation where a host is down, stuck or similar.
Cheers
#4 Updated by Rolandas Naujikas over 5 years ago
We had serious problem with delete in 4.12. VM was created and in the middle of starting - deleted. VM was still running, but opennebula thought resources are not in use. After new VM was created and same IP/MAC allocated. The solution to use shutdown(hard) don't work, because VM should be in the running state. We should probably add synchronous delete action for users.
#5 Updated by Ruben S. Montero over 5 years ago
- Status changed from Closed to Pending
- Target version changed from Release 4.14 to Release 5.0
Re-open to re-evalute it for next release.
#6 Updated by Carlos Martín over 5 years ago
- Related to Feature #3801: Revisit delete operation added
#7 Updated by Ruben S. Montero about 5 years ago
- Status changed from Pending to Closed
- Resolution changed from worksforme to fixed
This issue is now fixed, we have implemented a terminate action which acts as a sync delete operation on any final state. The previous delete operation is now under the recover interface for admins.