adamgoose
commented
7 years ago I've been seeing some of this going on as well. It looks like the volume driver is trying to delete disks that it doesn't recognize as volumes, namely my VM boot disks! >< I recently lost a very important volume due to this. I'm not sure if the command is being issued by Rancher, Docker, or this plugin.
5/29/2017 10:33:46 PMt=2017-05-30T03:33:46+0000 lvl=info msg="disk removed" disk=my-cesam elapsed=2.690921065s
5/29/2017 10:33:47 PMt=2017-05-30T03:33:47+0000 lvl=eror msg="request failed" error="googleapi: Error 400: The disk resource 'projects/bridgeconx/zones/us-central1-b/disks/neally-mysql' is already being used by 'projects/bridgeconx/zones/us-central1-b/instances/rancher-host-5', resourceInUseByAnotherResource"You can see the successful disk removal, followed by the failure. The failure is completely clogging the volume driver container logs. wtf!?
I'm currently using this with Rancher. With Rancher, I see log entries like:
I'm assuming the part
Unable to remove filesystem ...is from gce-docker but don't know for sure. Given I have a fleet of machines I'm using this for, its really hard to figure out what host / container / volume this is related to. If it included just the name of the GCE volume, that would solve my problem debugging what is going on.