Kubelet.cleanupOrphanedVolumes() compares list of volumes mounted to a node
with list of volumes that are required by pods scheduled on the node
("scheduled volume").
Both lists should contain real volumes, i.e. when a pod uses
PersistentVolumeClaim, the list must contain name of the bound volume instead
of name of the claim.
We do this because they will be recreated immediately by the
DaemonSet Controller. In addition, we also require a specific flag
(--ignore-daemonsets) when there are DaemonSet pods on the node.
Add a mutex to guard SetUpAt() and TearDownAt() calls - they should not
run in parallel. There is a race in these calls when there are two pods
using the same volume, one of them is dying and the other one starting.
TearDownAt() checks that a volume is not needed by any pods and detaches the
volume. It does so by counting how many times is the volume mounted
(GetMountRefs() call below).
When SetUpAt() of the starting pod already attached the volume and did not mount
it yet, TearDownAt() of the dying pod will detach it - GetMountRefs() does not
count with this volume.
These two threads run in parallel:
dying pod.TearDownAt("myVolume") starting pod.SetUpAt("myVolume")
| |
| AttachDisk("myVolume")
refs, err := mount.GetMountRefs() |
Unmount("myDir") |
if refs == 1 { |
| | Mount("myVolume", "myDir")
| | |
| DetachDisk("myVolume") |
| start containers - OOPS! The volume is detached!
|
finish the pod cleanup
Also, add some logs to cinder plugin for easier debugging in the future, add
a test and update the fake mounter to know about bind mounts.
Fixes#19860 (it may be easier to look at the issue to see exact sequence
to reproduce the bug and understand the fix).
When PersistentVolumeProvisionerController.reconcileClaim() is called with the
same claim in short succession (e.g. the claim is created by an user and
at the same time periodic check of all claims is scheduled), the second
reconcileClaim() call gets an old copy of the claim as its parameter.
The method should always reload the claim to get a fresh copy with all
annotations, possibly added by previous reconcileClaim() call.
The same applies to PersistentVolumeClaimBinder.syncClaim().
Also update all the test to store claims in "fake" API server before calling
syncClaim and reconcileClaim.
This change removes RuntimeCache in the pod workers and the syncPod() function.
Note that it doesn't deprecate RuntimeCache completely as other components
still rely on the cache.
