Skip to content

[ws-manager] Add event to workspace Pod when the PVC restored from the VolumeSnapshot #11367

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 5 commits into from
Jul 18, 2022
Merged

[ws-manager] Add event to workspace Pod when the PVC restored from the VolumeSnapshot #11367

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
e97df1a
Add event to the workspace pod related to the restoration of the PVC
jenting Jul 14, 2022
4dfd01c
Use the eventRecorder from manager object
jenting Jul 14, 2022
7db18e7
Use correct log variable to include OWI information in each log entry
jenting Jul 18, 2022
9d1a271
Update components/ws-manager/pkg/manager/manager.go
sagor999 Jul 18, 2022
737d7e1
Update components/ws-manager/pkg/manager/manager.go
sagor999 Jul 18, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
71 changes: 49 additions & 22 deletions components/ws-manager/pkg/manager/manager.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,9 +30,12 @@ import (
k8serr "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/runtime"
"k8s.io/apimachinery/pkg/types"
"k8s.io/apimachinery/pkg/util/wait"
"k8s.io/client-go/kubernetes"
covev1client "k8s.io/client-go/kubernetes/typed/core/v1"
"k8s.io/client-go/tools/record"
"k8s.io/client-go/util/retry"
"sigs.k8s.io/controller-runtime/pkg/client"

Expand Down Expand Up @@ -69,6 +72,8 @@ type Manager struct {

metrics *metrics

eventRecorder record.EventRecorder

api.UnimplementedWorkspaceManagerServer
regapi.UnimplementedSpecProviderServer
}
Expand Down Expand Up @@ -127,14 +132,19 @@ func New(config config.Configuration, client client.Client, rawClient kubernetes
return nil, err
}

broadcaster := record.NewBroadcaster()
broadcaster.StartRecordingToSink(&covev1client.EventSinkImpl{Interface: rawClient.CoreV1().Events("")})
eventRecorder := broadcaster.NewRecorder(runtime.NewScheme(), corev1.EventSource{Component: "ws-manager"})

m := &Manager{
Config: config,
Clientset: client,
RawClient: rawClient,
Content: cp,
clock: clock.System(),
subscribers: make(map[string]chan *api.SubscribeResponse),
wsdaemonPool: grpcpool.New(wsdaemonConnfactory, checkWSDaemonEndpoint(config.Namespace, client)),
Config: config,
Clientset: client,
RawClient: rawClient,
Content: cp,
clock: clock.System(),
subscribers: make(map[string]chan *api.SubscribeResponse),
wsdaemonPool: grpcpool.New(wsdaemonConnfactory, checkWSDaemonEndpoint(config.Namespace, client)),
eventRecorder: eventRecorder,
}
m.metrics = newMetrics(m)
m.OnChange = m.onChange
Expand Down Expand Up @@ -165,7 +175,7 @@ func (m *Manager) StartWorkspace(ctx context.Context, req *api.StartWorkspaceReq
clog.Info("StartWorkspace")
reqs, _ := protojson.Marshal(req)
safeReqs, _ := log.RedactJSON(reqs)
log.WithField("req", string(safeReqs)).Debug("StartWorkspace request received")
clog.WithField("req", string(safeReqs)).Debug("StartWorkspace request received")

// Make sure the objects we're about to create do not exist already
switch req.Type {
Expand Down Expand Up @@ -238,11 +248,11 @@ func (m *Manager) StartWorkspace(ctx context.Context, req *api.StartWorkspaceReq
// restore volume snapshot from handle
err = m.restoreVolumeSnapshotFromHandle(ctx, startContext.VolumeSnapshot.VolumeSnapshotName, startContext.VolumeSnapshot.VolumeSnapshotHandle)
if err != nil {
log.WithError(err).Error("was unable to restore volume snapshot")
clog.WithError(err).Error("was unable to restore volume snapshot")
return nil, err
}
} else if err != nil {
log.WithError(err).Error("was unable to get volume snapshot")
clog.WithError(err).Error("was unable to get volume snapshot")
return nil, err
}
}
Expand Down Expand Up @@ -310,20 +320,37 @@ func (m *Manager) StartWorkspace(ctx context.Context, req *api.StartWorkspaceReq
}

// we only calculate the time that PVC restoring from VolumeSnapshot
if createPVC && startContext.VolumeSnapshot != nil && startContext.VolumeSnapshot.VolumeSnapshotName != "" {
err := wait.PollWithContext(ctx, 100*time.Millisecond, 5*time.Minute, pvcRunning(m.Clientset, pvc.Name, pvc.Namespace))
if err == nil {
wsType := api.WorkspaceType_name[int32(req.Type)]
hist, err := m.metrics.volumeRestoreTimeHistVec.GetMetricWithLabelValues(wsType, req.Spec.Class)
if err != nil {
log.WithError(err).WithField("type", wsType).Warn("cannot get volume restore time histogram metric")
} else if endTime.IsZero() {
endTime = time.Now()
hist.Observe(endTime.Sub(startTime).Seconds())
}
if createPVC {
err = m.Clientset.Get(ctx, types.NamespacedName{Namespace: pod.Namespace, Name: pod.Name}, pod)
if err != nil {
return nil, xerrors.Errorf("unable to get workspace pod %s: %w", pod.Name, err)
}

log.WithError(err).Warn("unexpected error waiting for PVC")
err = wait.PollWithContext(ctx, 100*time.Millisecond, 5*time.Minute, pvcRunning(m.Clientset, pvc.Name, pvc.Namespace))
if err != nil {
if startContext.VolumeSnapshot != nil && startContext.VolumeSnapshot.VolumeSnapshotName != "" {
m.eventRecorder.Eventf(pod, corev1.EventTypeWarning, "PersistentVolumeClaim", "PVC %q restore from volume snapshot %q failed %v", pvc.Name, startContext.VolumeSnapshot.VolumeSnapshotName, err)
clog.WithError(err).Warnf("unexpected error waiting for PVC %s volume snapshot %s", pvc.Name, startContext.VolumeSnapshot.VolumeSnapshotName)
} else {
m.eventRecorder.Eventf(pod, corev1.EventTypeWarning, "PersistentVolumeClaim", "PVC %q create failed %v", pvc.Name, err)
clog.WithError(err).Warnf("unexpected error waiting for PVC %s", pvc.Name)
}
} else {
if startContext.VolumeSnapshot != nil && startContext.VolumeSnapshot.VolumeSnapshotName != "" {
m.eventRecorder.Eventf(pod, corev1.EventTypeNormal, "PersistentVolumeClaim", "PVC %q restored from volume snapshot %q successfully", pvc.Name, startContext.VolumeSnapshot.VolumeSnapshotName)

wsType := api.WorkspaceType_name[int32(req.Type)]
hist, err := m.metrics.volumeRestoreTimeHistVec.GetMetricWithLabelValues(wsType, req.Spec.Class)
if err != nil {
clog.WithError(err).WithField("type", wsType).Warn("cannot get volume restore time histogram metric")
} else if endTime.IsZero() {
endTime = time.Now()
hist.Observe(endTime.Sub(startTime).Seconds())
}
} else {
m.eventRecorder.Eventf(pod, corev1.EventTypeNormal, "PersistentVolumeClaim", "PVC %q created successfully", pvc.Name)
}
}
}

// remove annotation to signal that workspace pod was indeed created and scheduled on the node
Expand Down
8 changes: 1 addition & 7 deletions components/ws-manager/pkg/manager/monitor.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,10 +26,8 @@ import (
k8serr "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/runtime"
"k8s.io/apimachinery/pkg/types"
"k8s.io/apimachinery/pkg/watch"
covev1client "k8s.io/client-go/kubernetes/typed/core/v1"
"k8s.io/client-go/tools/record"
"sigs.k8s.io/controller-runtime/pkg/client"

Expand Down Expand Up @@ -94,10 +92,6 @@ type Monitor struct {

// CreateMonitor creates a new monitor
func (m *Manager) CreateMonitor() (*Monitor, error) {
broadcaster := record.NewBroadcaster()
broadcaster.StartRecordingToSink(&covev1client.EventSinkImpl{Interface: m.RawClient.CoreV1().Events("")})
eventRecorder := broadcaster.NewRecorder(runtime.NewScheme(), corev1.EventSource{Component: "ws-manager"})

monitorInterval := time.Duration(m.Config.HeartbeatInterval)
// Monitor interval is half the heartbeat interval to catch timed out workspaces in time.
// See https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampling_theorem why we need this.
Expand All @@ -117,7 +111,7 @@ func (m *Manager) CreateMonitor() (*Monitor, error) {

notifyPod: make(map[string]chan string),

eventRecorder: eventRecorder,
eventRecorder: m.eventRecorder,
}
res.eventpool = workpool.NewEventWorkerPool(res.handleEvent)
res.act = struct {
Expand Down