This PR adds the pod-level metrics for CPU and memory stats. cAdvisor
can get all pod cgroup information so we can add this pod-level CPU and
memory stats information from the corresponding pod cgroup
- Instead of using cm.capacity field to communicate device plugin resource
capacity, this PR changes to use an explicit cm.GetDevicePluginResourceCapacity()
function that returns device plugin resource capacity as well as any inactive
device plugin resource. Kubelet syncNodeStatus call this function during its
periodic run to update node status capacity and allocatable. After this call,
device plugin can remove the inactive device plugin resource from its allDevices
field as the update is already pushed to API server.
- Extends device plugin checkpoint data to record registered resources
so that we can finish resource removing even upon kubelet restarts.
- Passes sourcesReady from kubelet to device plugin to avoid removing
inactive pods during grace period of kubelet restart.
Automatic merge from submit-queue (batch tested with PRs 55839, 54495, 55884, 55983, 56069). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
deviceplugin: fix race when multiple plugins are registered
**What this PR does / why we need it**:
When registering multiple device plugins to Kubelet concurrently, there exists a race that crashes the Kubelet.
Consider two plugins: D1 and D2. The call order method is roughly
D1 -> manager.go:register -> endpoint.go:listAndWatch -> device_plugin_handler.go:(*D1).callback
D2 -> manager.go:register -> endpoint.go:listAndWatch -> device_plugin_handler.go:(*D2).callback
The callback function accesses HandlerImpl's allDevices map that maps (resourceName -> DeviceID). If both plugins reach these accesses at the same time, Kubelet crashes with "fatal error: concurrent map read and map write".
This can be solved by making sure handler is locked when allDevices are being updated. The functionality is needed to avoid Kubelet crashes when multiple device plugins are trying to register with Kubelet at the same moment. Occurs frequently when single binary tries to register itself as multiple plugins.
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 55841, 55948, 55945). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
CPU Manager: file state for all policies
**What this PR does / why we need it**:
Before this change, the new file-backed state was only enabled for the static CPU manager policy. This patch enables persistent state for all policies.
This PR fixes#55736 and the potential CPU resource leak described in that issue.
**Release note**:
```release-note
NONE
```
/kind bug
/sig node
/assign @balajismaniam
Automatic merge from submit-queue (batch tested with PRs 54837, 55970, 55912, 55898, 52977). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Improve kubelet cgroup
**What this PR does / why we need it**:
1.Use arg cgroupRoot,not nodeConfig.CgroupRoot
Using both arg cgroupRoot and nodeConfig.CgroupRoot is confused in function NewQOSContainerManager
2.improve cgroupmanager in qosContainerManager
3. improve arg "cgroupRoot" type in NewQOSContainerManager
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
```
- Changes the following KubeletConfiguration fields from `string` to
`map[string]string`:
- `EvictionHard`
- `EvictionSoft`
- `EvictionSoftGracePeriod`
- `EvictionMinimumReclaim`
- Adds flag parsing shims to maintain Kubelet's public flags API, while
enabling structured input in the file API.
- Also removes `kubeletconfig.ConfigurationMap`, which was an ad-hoc flag
parsing shim living in the kubeletconfig API group, and replaces it
with the `MapStringString` shim introduced in this PR. Flag parsing
shims belong in a common place, not in the kubeletconfig API.
I manually audited these to ensure that this wouldn't cause errors
parsing the command line for syntax that would have previously been
error free (`kubeletconfig.ConfigurationMap` was unique in that it
allowed keys to be provided on the CLI without values. I believe this was
done in `flags.ConfigurationMap` to facilitate the `--node-labels` flag,
which rightfully accepts value-free keys, and that this shim was then
just copied to `kubeletconfig`). Fortunately, the affected fields
(`ExperimentalQOSReserved`, `SystemReserved`, and `KubeReserved`) expect
non-empty strings in the values of the map, and as a result passing the
empty string is already an error. Thus requiring keys shouldn't break
anyone's scripts.
- Updates code and tests accordingly.
Regarding eviction operators, directionality is already implicit in the
signal type (for a given signal, the decision to evict will be made when
crossing the threshold from either above or below, never both). There is
no need to expose an operator, such as `<`, in the API. By changing
`EvictionHard` and `EvictionSoft` to `map[string]string`, this PR
simplifies the experience of working with these fields via the
`KubeletConfiguration` type. Again, flags stay the same.
Other things:
- There is another flag parsing shim, `flags.ConfigurationMap`, from the
shared flag utility. The `NodeLabels` field still uses
`flags.ConfigurationMap`. This PR moves the allocation of the
`map[string]string` for the `NodeLabels` field from
`AddKubeletConfigFlags` to the defaulter for the external
`KubeletConfiguration` type. Flags are layered on top of an internal
object that has undergone conversion from a defaulted external object,
which means that previously the mere registration of flags would have
overwritten any previously-defined defaults for `NodeLabels` (fortunately
there were none).
Automatic merge from submit-queue (batch tested with PRs 55764, 55683, 55468, 54409, 55546). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Enable file back state in static policy
**What this PR does / why we need it**:
Enables file back `State` in `static policy` and cpu manager + tests.
Upon policy start, state read from file is validated whether it meets the policy assumption. In case of any error, state is cleared.
Previous PR: #54408
Next PR: #54409
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Deviceplugin refactoring: merge func list and listwatch in endpoint into one
**What this PR does / why we need it**:
merge func list and listwatch in endpoint into one, since we won't call list func individually
**Which issue this PR fixes**
fixes#51993
Part2
**Special notes for your reviewer**:
/cc @jiayingz @RenaudWasTaken @vishh
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 55283, 55461, 55288, 53970, 55487). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Add more comments for DevicePluginHandlerImpl struct
**What this PR does / why we need it**:
Add more comments
**Special notes for your reviewer**:
@jiayingz PTAL.
**Release note**:
```
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Add admission handler for device resources allocation
**What this PR does / why we need it**:
Add admission handler for device resources allocation to fail fast during pod creation
**Which issue this PR fixes**
fixes#51592
**Special notes for your reviewer**:
@jiayingz Sorry, there is something wrong with my branch in #51895. And I think the existing comments in the PR might be too long for others to view. So I closed it and opened the new one, as we have basically reach an agreement on the implement :)
I have covered the functionality and unit test part here, and would set about the e2e part ASAP
/cc @jiayingz @vishh @RenaudWasTaken
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 54656, 54552, 54389, 53634, 54408). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Add file backed state to cpu manager
**What this PR does / why we need it**:
Adds file backed `State` implementation to cpu manger with tests.
Reads from `State` are done from memory, while each write triggers state save to a file.
Any failure in reading the state file results in empty state
Next PR: #54409
While moving device_plugin_handler_test.go from pkg/kubelet/cm/ to
pkg/kubelet/cm/deviceplugin/, we can no longer uses cm in its tests
because that would cause a cycle dependency. To solve this problem,
I moved the main cm GetResources functionality as well as part of the
current device plugin handler Allocate functionality into a new device
plugin handler function, GetDeviceRunContainerOptions(). This
refactoring is also needed by another PR 51895 that moves device
allocation into admission phase. Now device plugin handler Allocate()
first checks whether there is cached device runtime state and only
issues Allocate grpc call if there is no cached state available.
The new GetDeviceRunContainerOptions() function simply returns device
runtime config from the cached state. To support this change, extended the
podDevices struct and checkpoint data structure with device runtime state.
There is no use case known for passing artifacts per device as it currently exists. The current API is also
complex to use for simple clients. Hence this PR creates a flat namespace where artifacts like environment variables
and mount points apply globally to all devices returned as part of AllocateResponse proto.
Signed-off-by: Vishnu kannan <vishnuk@google.com>
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
fix typos: remove duplicated word in comments
**What this PR does / why we need it**: Remove the duplicated word `the` in comments
**Which issue this PR fixes** : fixes #
**Special notes for your reviewer**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 52959, 53790). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
remove redundancy code in setCPUCgroupConfig
fix#53925
Signed-off-by: yanxuean <yan.xuean@zte.com.cn>
**What this PR does / why we need it**:
The check of burstableCPUShares is redundancy. We have done it in MilliCPUToShares. It is responsibility of MilliCPUToShares.
```
func (m *qosContainerManagerImpl) setCPUCgroupConfig(configs map[v1.PodQOSClass]*CgroupConfig) error {
........
// set burstable shares based on current observe state
burstableCPUShares := MilliCPUToShares(burstablePodCPURequest)
if burstableCPUShares < uint64(MinShares) {
burstableCPUShares = uint64(MinShares)
}
```
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
Improveing code.
**Special notes for your reviewer**:
**Release note**:
```release-note
```
Automatic merge from submit-queue (batch tested with PRs 53119, 53753, 53795, 52981). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Reduce log spam in qos container manager
**What this PR does / why we need it**:
excessive log stmts make it hard to debug actual problems.
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 53297, 53328). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Cpu Manager - make CoreID's platform unique
**What this PR does / why we need it**:
Cpu Manager uses topology from cAdvisor(`/proc/cpuinfo`) where coreID's are socket unique - not platform unique - this causes problems on multi-socket platforms.
All code assumes unique coreID's (on platform) - `Discovery` function has been changed to assign CoreID as the lowest cpuID from all cpus belonging to the same core. This can be expressed as:
`CoreID=min(cpuID's on the same core)`
Since cpuID's are platform unique - above gives us guarantee that CoreID's will also be platform unique.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#53323
Automatic merge from submit-queue (batch tested with PRs 53525, 53652). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
apimachinery: remove ObjectCopier interface(s)
The big commit is a mechanical, transitive removal of the copier interfaces in all structs and function calls.
Automatic merge from submit-queue (batch tested with PRs 52662, 53547, 53588, 53573, 53599). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
In DevicePluginHandlerImpl.Allocate(), skips untracked extended resou…
…rces.
Otherwise, we would fail a Pod allocation request that has an extended
resource not managed by any device plugin.
**What this PR does / why we need it**:
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
https://github.com/kubernetes/kubernetes/issues/53548
**Special notes for your reviewer**:
**Release note**:
```release-note
Ignore extended resources that are not registered with kubelet
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Refactor nsenter
**What this PR does / why we need it**:
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#51273
**Special notes for your reviewer**:
/assign @jsafrane
**Release note**:
```release-note
None
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Eliminate extra CRI call during processing cpu set
**What this PR does / why we need it**:
Encountered this during `kubernetes/frakti` node e2e test.
When cpuset is not set, there's still plenty of `runtime.UpdateContainerResources` been called, which seems unnecessary.
cc @ConnorDoyle Make sense? Fixes: #53304
**Special notes for your reviewer**:
**Release note**:
```release-note
Only do UpdateContainerResources when cpuset is set
```
Automatic merge from submit-queue (batch tested with PRs 44596, 52708, 53163, 53167, 52692). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Fix the bad code comment and make the format unify
**What this PR does / why we need it**:
Fix the bad code comment and make the format unify
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
Make *fakeMountInterface in container_manager_unsupported_test.go implement mount.Interface again.
This was broken in #45724
**Release note**:
```release-note
NONE
```
/sig storage
/sig node
/cc @jsafrane, @vishh
Automatic merge from submit-queue (batch tested with PRs 51186, 50350, 51751, 51645, 51837)
Update Cadvisor Dependency
Fixes: https://github.com/kubernetes/kubernetes/issues/51832
This is the worst dependency update ever...
The root of the problem is the [name change of Sirupsen -> sirupsen](https://github.com/sirupsen/logrus/issues/570#issuecomment-313933276). This means that in order to update cadvisor, which venders the lowercase, we need to update all dependencies to use the lower-cased version. With that being said, this PR updates the following packages:
`github.com/docker/docker`
- `github.com/docker/distribution`
- `github.com/opencontainers/go-digest`
- `github.com/opencontainers/image-spec`
- `github.com/opencontainers/runtime-spec`
- `github.com/opencontainers/selinux`
- `github.com/opencontainers/runc`
- `github.com/mrunalp/fileutils`
- `golang.org/x/crypto`
- `golang.org/x/sys`
- `github.com/docker/go-connections`
- `github.com/docker/go-units`
- `github.com/docker/libnetwork`
- `github.com/docker/libtrust`
- `github.com/sirupsen/logrus`
- `github.com/vishvananda/netlink`
`github.com/google/cadvisor`
- `github.com/euank/go-kmsg-parser`
`github.com/json-iterator/go`
Fixed https://github.com/kubernetes/kubernetes/issues/51832
```release-note
Fix journalctl leak on kubelet restart
Fix container memory rss
Add hugepages monitoring support
Fix incorrect CPU usage metrics with 4.7 kernel
Add tmpfs monitoring support
```
Automatic merge from submit-queue (batch tested with PRs 50072, 51744)
Deviceplugin checkpoint
**What this PR does / why we need it**:
Extends on top of PR 51209 to checkpoint device to pod allocation information on Kubelet to recover from Kubelet restarts.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
```
Automatic merge from submit-queue
Make /var/lib/kubelet as shared during startup
This is part of ~~https://github.com/kubernetes/community/pull/589~~https://github.com/kubernetes/community/pull/659
We'd like kubelet to be able to consume mounts from containers in the future, therefore kubelet should make sure that `/var/lib/kubelet` has shared mount propagation to be able to see these mounts.
On most distros, root directory is already mounted with shared mount propagation and this code will not do anything. On older distros such as Debian Wheezy, this code detects that `/var/lib/kubelet` is a directory on `/` which has private mount propagation and kubelet bind-mounts `/var/lib/kubelet` as rshared.
Both "regular" linux mounter and `NsenterMounter` are updated here.
@kubernetes/sig-storage-pr-reviews @kubernetes/sig-node-pr-reviews
@vishh
Release note:
```release-note
Kubelet re-binds /var/lib/kubelet directory with rshared mount propagation during startup if it is not shared yet.
```
Kubelet makes sure that /var/lib/kubelet is rshared when it starts.
If not, it bind-mounts it with rshared propagation to containers
that mount volumes to /var/lib/kubelet can benefit from mount propagation.
Automatic merge from submit-queue (batch tested with PRs 51391, 51338, 51340, 50773, 49599)
Delete "hugetlb" from whitelistControllers
**What this PR does / why we need it**:
Delete "hugetlb" from whitelistControllers
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#50770
**Special notes for your reviewer**:
**Release note**:
```
NONE
```
