Starting replicator/reconstructor with *_workers and no devices in ring (yet) runs inline

Bug #1830074 reported by clayg
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
OpenStack Object Storage (swift)
New
Undecided
Unassigned

Bug Description

I've had reports that starting a replicator/re-constructor after "adding a new node" ignores the workers option.

The code looks for locals devices to divvy them up between the available workers. But if there's no devices it returns an empty set and falls back to inline mode [1].

Once the deamon goes into inline mode, if new rings come down with devices it just restarts the run loop - it doesn't bounce out to the daemon strategy to redeploy workers.

1. https://github.com/openstack/swift/blob/b4c8cf192a39004645e2608d4cfc0ac5dcf3b516/swift/common/daemon.py#L224

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.