Skip to content
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

Enhancement freeze logic #180

Merged
merged 5 commits into from
Oct 31, 2018
Merged

Enhancement freeze logic #180

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
e5c419b
(Un)freezing of machine controller based on APIServer availability
prashanth26 Oct 15, 2018
5c751f5
Testing Framework now supports faking of object tracker returns/delays
prashanth26 Oct 15, 2018
adef455
Added test cases for machine controller freezing
prashanth26 Oct 22, 2018
53ad7ff
Minor fixes to enhance machine controller freeze
prashanth26 Oct 23, 2018
bdf03bf
Get dummy_name objects instead of blank
prashanth26 Oct 31, 2018
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
2 changes: 2 additions & 0 deletions Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,8 @@ start:
--machine-set-scale-timeout=20 \
--machine-safety-orphan-vms-period=30 \
--machine-safety-overshooting-period=1 \
--machine-safety-apiserver-period=1m \
prashanth26 marked this conversation as resolved.
Show resolved Hide resolved
--machine-safety-apiserver-timeout=30s \
--v=2

#################################################################
Expand Down
19 changes: 12 additions & 7 deletions cmd/machine-controller-manager/app/options/options.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,13 +61,15 @@ func NewMCMServer() *MCMServer {
LeaderElection: leaderelectionconfig.DefaultLeaderElectionConfiguration(),
ControllerStartInterval: metav1.Duration{Duration: 0 * time.Second},
SafetyOptions: machineconfig.SafetyOptions{
SafetyUp: 2,
SafetyDown: 1,
MachineHealthTimeout: 10,
MachineDrainTimeout: 5,
MachineSetScaleTimeout: 20,
MachineSafetyOrphanVMsPeriod: 30,
MachineSafetyOvershootingPeriod: 1,
SafetyUp: 2,
SafetyDown: 1,
MachineHealthTimeout: 10,
MachineDrainTimeout: 5,
MachineSetScaleTimeout: 20,
MachineSafetyOrphanVMsPeriod: 30,
MachineSafetyOvershootingPeriod: 1,
MachineSafetyAPIServerStatusPeriod: metav1.Duration{Duration: 1 * time.Minute},
MachineSafetyAPIServerStatusTimeout: metav1.Duration{Duration: 30 * time.Second},
},
},
}
Expand Down Expand Up @@ -100,6 +102,9 @@ func (s *MCMServer) AddFlags(fs *pflag.FlagSet) {
fs.Int32Var(&s.SafetyOptions.MachineSafetyOrphanVMsPeriod, "machine-safety-orphan-vms-period", s.SafetyOptions.MachineSafetyOrphanVMsPeriod, "Time period (in minutes) used to poll for orphan VMs by safety controller.")
fs.Int32Var(&s.SafetyOptions.MachineSafetyOvershootingPeriod, "machine-safety-overshooting-period", s.SafetyOptions.MachineSafetyOvershootingPeriod, "Time period (in minutes) used to poll for overshooting of machine objects backing a machineSet by safety controller.")

fs.DurationVar(&s.SafetyOptions.MachineSafetyAPIServerStatusTimeout.Duration, "machine-safety-apiserver-timeout", s.SafetyOptions.MachineSafetyAPIServerStatusTimeout.Duration, "Timeout (in duration) for which the APIServer can be down before declare the machine controller frozen by safety controller")
fs.DurationVar(&s.SafetyOptions.MachineSafetyAPIServerStatusPeriod.Duration, "machine-safety-apiserver-period", s.SafetyOptions.MachineSafetyAPIServerStatusPeriod.Duration, "Period (in duration) used to poll for APIServer's health by safety controller")
prashanth26 marked this conversation as resolved.
Show resolved Hide resolved

leaderelectionconfig.BindFlags(&s.LeaderElection, fs)
// TODO: DefaultFeatureGate is global and it adds all k8s flags
// utilfeature.DefaultFeatureGate.AddFlag(fs)
Expand Down
5 changes: 5 additions & 0 deletions pkg/controller/controller.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,6 +98,7 @@ func NewController(
machineDeploymentQueue: workqueue.NewNamedRateLimitingQueue(workqueue.DefaultControllerRateLimiter(), "machinedeployment"),
machineSafetyOrphanVMsQueue: workqueue.NewNamedRateLimitingQueue(workqueue.DefaultControllerRateLimiter(), "machinesafetyorphanvms"),
machineSafetyOvershootingQueue: workqueue.NewNamedRateLimitingQueue(workqueue.DefaultControllerRateLimiter(), "machinesafetyovershooting"),
machineSafetyAPIServerQueue: workqueue.NewNamedRateLimitingQueue(workqueue.DefaultControllerRateLimiter(), "machinesafetyapiserver"),
safetyOptions: safetyOptions,
}

Expand Down Expand Up @@ -334,6 +335,7 @@ func NewController(
// running of different safety loop on MCM startup.
controller.machineSafetyOrphanVMsQueue.Add("")
controller.machineSafetyOvershootingQueue.Add("")
controller.machineSafetyAPIServerQueue.Add("")

machineInformer.Informer().AddEventHandler(cache.ResourceEventHandlerFuncs{
// addMachineToSafety makes sure machine objects does not overshoot
Expand Down Expand Up @@ -397,6 +399,7 @@ type controller struct {
machineDeploymentQueue workqueue.RateLimitingInterface
machineSafetyOrphanVMsQueue workqueue.RateLimitingInterface
machineSafetyOvershootingQueue workqueue.RateLimitingInterface
machineSafetyAPIServerQueue workqueue.RateLimitingInterface
// syncs
secretSynced cache.InformerSynced
nodeSynced cache.InformerSynced
Expand Down Expand Up @@ -429,6 +432,7 @@ func (c *controller) Run(workers int, stopCh <-chan struct{}) {
defer c.machineDeploymentQueue.ShutDown()
defer c.machineSafetyOrphanVMsQueue.ShutDown()
defer c.machineSafetyOvershootingQueue.ShutDown()
defer c.machineSafetyAPIServerQueue.ShutDown()

if !cache.WaitForCacheSync(stopCh, c.secretSynced, c.nodeSynced, c.openStackMachineClassSynced, c.awsMachineClassSynced, c.azureMachineClassSynced, c.gcpMachineClassSynced, c.alicloudMachineClassSynced, c.machineSynced, c.machineSetSynced, c.machineDeploymentSynced) {
runtimeutil.HandleError(fmt.Errorf("Timed out waiting for caches to sync"))
Expand Down Expand Up @@ -457,6 +461,7 @@ func (c *controller) Run(workers int, stopCh <-chan struct{}) {
createWorker(c.machineDeploymentQueue, "ClusterMachineDeployment", maxRetries, true, c.reconcileClusterMachineDeployment, stopCh, &waitGroup)
createWorker(c.machineSafetyOrphanVMsQueue, "ClusterMachineSafetyOrphanVMs", maxRetries, true, c.reconcileClusterMachineSafetyOrphanVMs, stopCh, &waitGroup)
createWorker(c.machineSafetyOvershootingQueue, "ClusterMachineSafetyOvershooting", maxRetries, true, c.reconcileClusterMachineSafetyOvershooting, stopCh, &waitGroup)
createWorker(c.machineSafetyAPIServerQueue, "ClusterMachineAPIServer", maxRetries, true, c.reconcileClusterMachineSafetyAPIServer, stopCh, &waitGroup)
}

<-stopCh
Expand Down
Loading