[
https://issues.apache.org/jira/browse/MESOS-621?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13732052#comment-13732052
]
Thomas Marshall commented on MESOS-621:
---------------------------------------
I don't see how this is a bug. The current semantics are that
resourcesRecovered or resourcesUnused get called for all resources that get
allocated, so the resources on the removed slave will be unallocated from the
framework when that happens (see, for example, Master::offer, where we call
resourcesRecovered if the offered resources are on a removed slave).
You could argue that these semantics don't make much sense, since a framework
can have resources allocated to it that don't actually exist in the cluster
anymore, and we can certainly change them, but I believe everything should work
as intended with the code the way it is right now.
Its also worth noting that removing frameworks uses the same semantics (see,
again, Master::offer where we call resourcesRecovered if the framework
receiving the offer has been removed).
> HierarchicalAllocator::slaveRemoved doesn't properly handle framework
> allocations/resources
> -------------------------------------------------------------------------------------------
>
> Key: MESOS-621
> URL: https://issues.apache.org/jira/browse/MESOS-621
> Project: Mesos
> Issue Type: Bug
> Reporter: Vinod Kone
> Assignee: Thomas Marshall
> Fix For: 0.14.0
>
>
> Currently a slaveRemoved() simply removes the slave from 'slaves' map and
> slave's resources from 'roleSorter'. Looking at resourcesRecovered(), more
> things need to be done when a slave is removed (e.g., framework
> unallocations).
> It would be nice to fix this and have a test for this.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
