Move back to FailureThreshold failures of /gshealthz #3160
Conversation
|
Analysis of a recent failure. |
|
Build Failed 😱 Build Id: f50aa99d-ed0a-458e-bca8-37f253bebb0b To get permission to view the Cloud Build view, join the agones-discuss Google Group. |
| // | ||
| // Note that in general, FailureThreshold could also be infinite - the controller | ||
| // and sidecar are responsible for health management. | ||
| FailureThreshold: gs.Spec.Health.FailureThreshold, |
There was a problem hiding this comment.
The default value is 3, right?
There was a problem hiding this comment.
Yeah, and can be raised, in theory.
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: gongmax, zmerlynn The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
New changes are detected. LGTM label has been removed. |
|
Build Failed 😱 Build Id: 64d0d412-7220-4b1f-8d97-6e0bef8f770c To get permission to view the Cloud Build view, join the agones-discuss Google Group. |
zmerlynn
force-pushed
the
bump-healthcheck-window-back
branch
from
14574eb to
1e935d1
Compare
This again reverts part of googleforgames#3072. In looking at recent failures, in cases where the sidecar is slow to come up due to networking issues, /gshealthz fails, resulting often in the unnecessary restart of the game server. This returns to the more generous failure threshold from prior to to just make this ~infinite, i.e. neuter kubelet liveness, since health is really owned by the controller and sidecar.
zmerlynn
force-pushed
the
bump-healthcheck-window-back
branch
from
1e935d1 to
bdb648b
Compare
|
Build Succeeded 👏 Build Id: 9fb4ad91-f23c-4bb4-9c12-91e08b52b32b The following development artifacts have been built, and will exist for the next 30 days:
A preview of the website (the last 30 builds are retained): To install this version:
|
This again reverts part of #3072. In looking at recent failures, in cases where the sidecar is slow to come up due to networking issues,
/gshealthzfails, resulting often in the unnecessary restart of the game server.This returns to the more generous failure threshold from prior to to just make this ~infinite, i.e. neuter kubelet liveness, since health is really owned by the controller and sidecar.