Because the async resource cleanup doesn't check whether the channel it's attempting to send the delete commands over is currently valid, it means that it will keep attempting to delete the directory over and over. This will cause the GC to hold onto the Channel/machine resources.
In installations that use the Jenkins cloud feature, nodes that were deleted shortly/immediately after a job being finished will slowly leak resources till they OOM.
|Field||Original Value||New Value|
|Assignee||Oliver Gondža [ olivergondza ]||Matthew Mitchell [ mmitche ]|
|Status||Open [ 1 ]||In Progress [ 3 ]|
|Status||In Progress [ 3 ]||In Review [ 10005 ]|
|Resolution||Fixed [ 1 ]|
|Status||In Review [ 10005 ]||Resolved [ 5 ]|
|Status||Resolved [ 5 ]||Fixed but Unreleased [ 10203 ]|
|Status||Fixed but Unreleased [ 10203 ]||Closed [ 6 ]|