Restart all pods in a deployment (rolling restart)

Learning Kubernetes · 4:24

Listen on 93

Lyrics

[Verse 1]
When your pods are feeling tired and old
And updates just won't take hold
There's a magic command that works so well
Rolling restart, I'll tell you how to tell

Type kubectl rollout restart
Then deployment slash the name
Every pod will cycle through
Zero downtime in this game

[Chorus]
Rolling restart, rolling restart
Keep your service running smart
One by one they'll cycle through
Rolling restart will see you through
Max surge up and max unavailable
Keep your limits reasonable
Rolling restart, that's the way
To refresh without delay

[Verse 2]
Don't forget the record flag
For history you can track and tag
Revision history limit set
Ten old sets you'll keep, don't fret

Rolling update strategy
Max surge and max unavailable
Control the flow of pods that go
Keep your service always reachable

[Chorus]
Rolling restart, rolling restart
Keep your service running smart
One by one they'll cycle through
Rolling restart will see you through
Max surge up and max unavailable
Keep your limits reasonable
Rolling restart, that's the way
To refresh without delay

[Bridge]
StatefulSets and DaemonSets
Jobs and pods, no regrets
Rolling restarts work for all
Kubernetes will heed your call

Annotate for better tracking
Never leave your history lacking
Rollout restart is the command
That keeps your cluster running grand

[Chorus]
Rolling restart, rolling restart
Keep your service running smart
One by one they'll cycle through
Rolling restart will see you through
Max surge up and max unavailable
Keep your limits reasonable
Rolling restart, that's the way
To refresh without delay

[Outro]
Kubectl rollout restart deployment
That's your faithful friend
Zero downtime, pods refreshed
On this you can depend

← Pause/resume a rollout (batch multiple changes) | StatefulSets (databases, stateful apps — ordered, stable identity) →