This change is prerequisite for implementing iSCSI attacher
and detacher.
In order to use chap authentication at iSCSI plugin after
implementing attacher and detacher, secret is needed at
AttachDisk() which is called from WaitForAttach().
To obtain secret, pod information is required, but
WaitForAttach() doesn't pass pod information inside.
This patch adds 'pod' as an argument of WaitForAttach()
and adds changes to drivers who implements WaitForAttach().
Fixes#48953
Automatic merge from submit-queue (batch tested with PRs 42369, 42375, 42397, 42435, 42455)
[Bug Fix]: Avoid evicting more pods than necessary by adding Timestamps for fsstats and ignoring stale stats
Continuation of #33121. Credit for most of this goes to @sjenning. I added volume fs timestamps.
**why is this a bug**
This PR attempts to fix part of https://github.com/kubernetes/kubernetes/issues/31362 which results in multiple pods getting evicted unnecessarily whenever the node runs into resource pressure. This PR reduces the chances of such disruptions by avoiding reacting to old/stale metrics.
Without this PR, kubernetes nodes under resource pressure will cause unnecessary disruptions to user workloads.
This PR will also help deflake a node e2e test suite.
The eviction manager currently avoids evicting pods if metrics are old. However, timestamp data is not available for filesystem data, and this causes lots of extra evictions.
See the [inode eviction test flakes](https://k8s-testgrid.appspot.com/google-node#kubelet-flaky-gce-e2e) for examples.
This should probably be treated as a bugfix, as it should help mitigate extra evictions.
cc: @kubernetes/sig-storage-pr-reviews @kubernetes/sig-node-pr-reviews @vishh @derekwaynecarr @sjenning
This implements Bulk volume polling using ideas presented by
justin in https://github.com/kubernetes/kubernetes/pull/39564
But it changes the implementation to use an interface
and doesn't affect other implementations.
Automatic merge from submit-queue
Remove unused WaitForDetach from Detacher interface and plugins
See issue #33128 and PR #33270
We can't rely on the device name provided by OpenStack Cinder, and thus
must perform detection based on the drive serial number (aka It's cinder ID)
on the kubelet itself.
This needs to be removed now, as part of #33128, as the code can't be
updated to attempt device detection and fallback through to the Cinder
provided deviceName, as detection "fails" when the device is gone, and
if cinder has reported a deviceName that another volume has used in
relaity, then this will block forever (or until the other, unreleated,
volume has been detached)
Automatic merge from submit-queue
Initial work on running windows containers on Kubernetes
<!-- Thanks for sending a pull request! Here are some tips for you:
1. If this is your first time, read our contributor guidelines https://github.com/kubernetes/kubernetes/blob/master/CONTRIBUTING.md and developer guide https://github.com/kubernetes/kubernetes/blob/master/docs/devel/development.md
2. If you want *faster* PR reviews, read how: https://github.com/kubernetes/kubernetes/blob/master/docs/devel/faster_reviews.md
3. Follow the instructions for writing a release note: https://github.com/kubernetes/kubernetes/blob/master/docs/devel/pull-requests.md#release-notes
-->
This is the first stab at getting the Kubelet running on Windows (fixes#30279), and getting it to deploy network-accessible pods that consist of Windows containers. Thanks @csrwng, @jbhurat for helping out.
The main challenge with Windows containers at this point is that container networking is not supported. In other words, each container in the pod will get it's own IP address. For this reason, we had to make a couple of changes to the kubelet when it comes to setting the pod's IP in the Pod Status. Instead of using the infra-container's IP, we use the IP address of the first container.
Other approaches we investigated involved "disabling" the infra container, either conditionally on `runtime.GOOS` or having a separate windows-docker container runtime that re-implemented some of the methods (would require some refactoring to avoid maintainability nightmare).
Other changes:
- The default docker endpoint was removed. This results in the docker client using the default for the specific underlying OS.
More detailed documentation on how to setup the Windows kubelet can be found at https://docs.google.com/document/d/1IjwqpwuRdwcuWXuPSxP-uIz0eoJNfAJ9MWwfY20uH3Q.
cc: @ikester @brendandburns @jstarks
This has been unused since 542f2dc7, and relies on deviceName, which
can no longer be relied upon (see issue #33128).
This needs to be removed now, as part of #33128, as the code can't be
updated to attempt device detection and fallback through to the Cinder
provided deviceName, as detection "fails" when the device is gone, and
if cinder has reported a deviceName that another volume has used in
relaity, then this will block forever (or until the other, unreleated,
volume has been detached)
At master volume reconciler, the information about which volumes are
attached to nodes is cached in actual state of world. However, this
information might be out of date in case that node is terminated (volume
is detached automatically). In this situation, reconciler assume volume
is still attached and will not issue attach operation when node comes
back. Pods created on those nodes will fail to mount.
This PR adds the logic to periodically sync up the truth for attached volumes kept in the actual state cache. If the volume is no longer attached to the node, the actual state will be updated to reflect the truth. In turn, reconciler will take actions if needed.
To avoid issuing many concurrent operations on cloud provider, this PR
tries to add batch operation to check whether a list of volumes are
attached to the node instead of one request per volume.
More details are explained in PR #33760
We had another bug where we confused the hostname with the NodeName.
To avoid this happening again, and to make the code more
self-documenting, we use types.NodeName (a typedef alias for string)
whenever we are referring to the Node.Name.
A tedious but mechanical commit therefore, to change all uses of the
node name to use types.NodeName
Also clean up some of the (many) places where the NodeName is referred
to as a hostname (not true on AWS), or an instanceID (not true on GCE),
etc.
Modify attach/detach controller to keep track of volumes to report
attached in Node VolumeToAttach status.
Modify kubelet volume manager to wait for volume to show up in Node
VolumeToAttach status.
Implement exponential backoff for errors in volume manager and attach
detach controller
This commit adds a new volume manager in kubelet that synchronizes
volume mount/unmount (and attach/detach, if attach/detach controller
is not enabled).
This eliminates the race conditions between the pod creation loop
and the orphaned volumes loops. It also removes the unmount/detach
from the `syncPod()` path so volume clean up never blocks the
`syncPod` loop.
Automatic merge from submit-queue
Add support for PersistentVolumeClaim in Attacher/Detacher interface
The attach detach interface does not support volumes which are referenced through PVCs. This PR adds that support
Automatic merge from submit-queue
Refactor persistent volume controller
Here is complete persistent controller as designed in https://github.com/pmorie/pv-haxxz/blob/master/controller.go
It's feature complete and compatible with current binder/recycler/provisioner. No new features, it *should* be much more stable and predictable.
Testing
--
The unit test framework is quite complicated, still it was necessary to reach reasonable coverage (78% in `persistentvolume_controller.go`). The untested part are error cases, which are quite hard to test in reasonable way - sure, I can inject a VersionConflictError on any object update and check the error bubbles up to appropriate places, but the real test would be to run `syncClaim`/`syncVolume` again and check it recovers appropriately from the error in the next periodic sync. That's the hard part.
Organization
---
The PR starts with `rm -rf kubernetes/pkg/controller/persistentvolume`. I find it easier to read when I see only the new controller without old pieces scattered around.
[`types.go` from the old controller is reused to speed up matching a bit, the code looks solid and has 95% unit test coverage].
I tried to split the PR into smaller patches, let me know what you think.
~~TODO~~
--
* ~~Missing: provisioning, recycling~~.
* ~~Fix integration tests~~
* ~~Fix e2e tests~~
@kubernetes/sig-storage
<!-- Reviewable:start -->
---
This change is [<img src="http://reviewable.k8s.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](http://reviewable.k8s.io/reviews/kubernetes/kubernetes/24331)
<!-- Reviewable:end -->
Fixes#15632
- Expand Attacher/Detacher interfaces to break up work more
explicitly.
- Add arguments to all functions to avoid having implementers store
the data needed for operations.
- Expand unit tests to check that Attach, Detach, WaitForAttach,
WaitForDetach, MountDevice, and UnmountDevice get call where
appropriet.
Jan Safranek
