-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
4.16.z
-
Quality / Stability / Reliability
-
False
-
-
None
-
None
-
None
-
None
-
None
-
CORENET Sprint 280
-
1
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
azure-vips nat is left empty even when the service is back up
Version-Release number of selected component (if applicable):
OCP 4.16.42
How reproducible:
requires precise timing and should not be super frequent
Steps to Reproduce:
1. need the api server watcher to create and remove files:
08:46:40.337386 3183 run.go:160] healthcheck failed, created downfile 10.20.30.40.down 08:49:26.410244 3183 run.go:174] healthcheck succeeded, removed downfile 10.20.30.40.down
2.
need a systemd reload around the same time. This seems to happen following a mco update
08:49:07 rpm-ostree[631231]: Reading config file '/etc/rpm-ostreed.conf'
..
08:49:17 systemd[1]: Reloading.
08:49:18 systemd[1]: Reloading.
08:49:20 systemd[1]: Reloading.
08:49:22 systemd[1]: Reloading.
08:49:24 systemd[1]: Reloading.
08:49:26 systemd[1]: Reloading.
08:49:27 systemd[1]: Reloading.
08:49:28 systemd[1]: Reloading.
08:49:29 systemd[1]: Reloading.
08:49:31 systemd[1]: Reloading.
3.
Actual results:
azure-vips table left empty
nft list chain ip nat azure-vips
table ip nat {
chain azure-vips {
}
}
Expected results:
a timer should have triggered to reconcile the azure-vips table with content of /var/run/cloud-routes/
Additional info:
