Closed olix0r closed 7 hours ago
Reduces start time from 20s to 4s:
On edge:
:; time (kubectl rollout restart -n linkerd deploy linkerd-destination && kubectl wait --for=condition=ready pod -l linkerd.io/control-plane-component=des
tination -n linkerd --timeout=600s)
deployment.apps/linkerd-destination restarted
pod/linkerd-destination-7587dbfcbf-5k79q condition met
pod/linkerd-destination-5774cc74d7-rphgl condition met
( kubectl rollout restart -n linkerd deploy linkerd-destination && kubectl ) 0.23s user 0.03s system 1% cpu 20.454 total
After:
:; time (kubectl rollout restart -n linkerd deploy linkerd-destination && kubectl wait --for=condition=ready pod -l linkerd.io/control-plane-component=destination -n linkerd
--timeout=600s)
deployment.apps/linkerd-destination restarted
pod/linkerd-destination-5c99cb9657-q877q condition met
pod/linkerd-destination-848f56c669-gpjfk condition met
( kubectl rollout restart -n linkerd deploy linkerd-destination && kubectl ) 0.24s user 0.04s system 7% cpu 3.587 total
The policy container is configured differently than all other controllers: other controllers configure an
initialDelaySeconds
on theirlivenessProbe
but not on theirreadinessProbe
. The policy container, however, includes this configuration on itsreadinessProbe
but not on itslivenessProbe
.This commit fixes the policy container to match the other controllers.
This reduces pod readiness time from 20s to 4s.