A number of race conditions exist when pods are terminated early in
their lifecycle because components in the kubelet need to know "no
running containers" or "containers can't be started from now on" but
were relying on outdated state.
Only the pod worker knows whether containers are being started for
a given pod, which is required to know when a pod is "terminated"
(no running containers, none coming). Move that responsibility and
podKiller function into the pod workers, and have everything that
was killing the pod go into the UpdatePod loop. Split syncPod into
three phases - setup, terminate containers, and cleanup pod - and
have transitions between those methods be visible to other
components. After this change, to kill a pod you tell the pod worker
to UpdatePod({UpdateType: SyncPodKill, Pod: pod}).
Several places in the kubelet were incorrect about whether they
were handling terminating (should stop running, might have
containers) or terminated (no running containers) pods. The pod worker
exposes methods that allow other loops to know when to set up or tear
down resources based on the state of the pod - these methods remove
the possibility of race conditions by ensuring a single component is
responsible for knowing each pod's allowed state and other components
simply delegate to checking whether they are in the window by UID.
Removing containers now no longer blocks final pod deletion in the
API server and are handled as background cleanup. Node shutdown
no longer marks pods as failed as they can be restarted in the
next step.
See https://docs.google.com/document/d/1Pic5TPntdJnYfIpBeZndDelM-AbS4FN9H2GTLFhoJ04/edit# for details
podVolumesExist() should consider also uncertain volumes (where kubelet
does not know if a volume was fully unmounted) when checking for pod's
volumes. Added GetPossiblyMountedVolumesForPod for that.
Adding uncertain mounts to GetMountedVolumesForPod would potentially break
other callers (e.g. `verifyVolumesMountedFunc`).
Volumes that are provisioned with `VolumeMode: Block` often have a
MetrucsProvider interface declared in their type. However, the
MetricsProvider should implement a GetMetrics() function. In the cases
where the storage drivers do not implement GetMetrics(), a panic can
occur.
Usual type-assertions are not sufficient in this case. All assertions
assume the interface is present. There is no straight forward way to
verify that a valid GetMetrics() function is provided.
By adding SupportsMetrics(), storage driver implementations require
careful reviewing for metrics support.
check in-memory cache whether volumes are still mounted and check disk directory for the volume paths instead of mounted volumes check
Signed-off-by: Mucahit Kurt <mucahitkurt@gmail.com>
- Rename MapDevice to MapPodDevice in BlockVolumeMapper
- Add UnmapPodDevice in BlockVolumeUnmapper (This will be used by csi driver later)
- Add CustomBlockVolumeMapper and CustomBlockVolumeUnmapper interface
- Move SetUpDevice and MapPodDevice to CustomBlockVolumeMapper
- Move TearDownDevice and UnmapPodDevice to CustomBlockVolumeUnmapper
- Implement CustomBlockVolumeMapper only in local and csi plugin
- Implement CustomBlockVolumeUnmapper only in fc, iscsi, rbd, and csi plugin
- Change MapPodDevice to return path and SetUpDevice not to return path
DesiredStateOfWorldPopulator should skip a volume that is not used in any
pod. "Used" means either mounted (via volumeMounts) or used as raw block
device (via volumeDevices).
Especially when block feature is disabled, a block volume must not get into
DesiredStateOfWorld, because it would be formatted and mounted there.
Automatic merge from submit-queue (batch tested with PRs 37228, 40146, 40075, 38789, 40189)
Cleanup temp dirs
So funny story my /tmp ran out of space running the unit tests so I am cleaning up all the temp dirs we create.
