Bug 1861455: Remove initial haproxy template commitAndReload #165
Conversation
openshift-ci-robot
added
the
do-not-merge/work-in-progress
openshift-ci-robot
added
the
approved
sgreene570
force-pushed
the
router-remove-initial-sync
branch
2 times, most recently
from
7aafc41 to
6ca2295
Compare
sgreene570
changed the title
openshift-ci-robot
removed
the
do-not-merge/work-in-progress
|
@sgreene570: This pull request references Bugzilla bug 1861455, which is valid. The bug has been moved to the POST state. The bug has been updated to refer to the pull request using the external bug tracker. 3 validation(s) were run on this bug
DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
openshift-ci-robot
added
bugzilla/severity-urgent
sgreene570
force-pushed
the
router-remove-initial-sync
branch
from
6ca2295 to
5c8ec99
Compare
If a router reload fails after a router pod becomes ready, we need a way to alert cluster admins that newly created route resources are not being applied to the cluster.
The first call to `commitAndReload` bypasses the rate limited reload logic that also takes into account route sync status. Removing this initial router reload call will prevent the router from starting in a "routeless" state: that is, a state were the router is running before it has begun watching route resources.
sgreene570
force-pushed
the
router-remove-initial-sync
branch
from
5c8ec99 to
becfcb0
Compare
|
Thanks! This looks terrific! |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: Miciah, sgreene570 The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
4 similar comments
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
@sgreene570: All pull requests linked via external trackers have merged: openshift/router#165. Bugzilla bug 1861455 has been moved to the MODIFIED state. DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
/cherry-pick release-4.5 |
|
@sgreene570: #165 failed to apply on top of branch "release-4.5": DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
6 participants
Waiting for the router factory/controller to call the router plugin's
Commitfunction guarantees that route/endpoint resources are available to the template code, and thus removes the possibility of a "routeless" router. A "routeless" router can be detrimental during upgrades if the routes carrying over during the upgrade break a newer version HAProxy, since the "routeless" configuration will continue to run if future successive reloads fail, and the upgrade will succeed when it should not.The initial (premature) call to the router's
commitAndReload()function outside of the rate limited loop is older code (3.11 era) and should be removed.This PR also adds a failed reload count metric for tracking failed reloads that happen well after a new router pod is created. This will be used for cluster alerting.
See the linked BZ for more context.
/assign @Miciah
/cc @frobware @danehans