jiocloud-devops team mailing list archive
-
jiocloud-devops team
-
Mailing list archive
-
Message #00069
[Bug 1449422] Re: vrouter service went bad on a particular node
The vRouter issue has gone away since we've enabled the collector,
however the collector seems to put a significant strain on the
controller. Easily consuming 10 full CPU cores.
anshup [1:41 PM]
basically we enabled collector earlier as a solution to vrouter issue -
but the collector itself causes heavier load / time outs / other issues in the existing setup -
(Thats why it was disabled earlier as well I believe)
so after we faced one of such collector caused issues again after enabling it 3 weeks back for vrouter issue
collector was disabled to avoid the load related issue
mayankkapoor: Creating a separate bug for collector load issue here
https://bugs.launchpad.net/jio/+bug/1465259.
--
You received this bug notification because you are a member of Reliance
Jio DevOps, which is a bug assignee.
https://bugs.launchpad.net/bugs/1449422
Title:
vrouter service went bad on a particular node
Status in Jio:
Fix Released
Bug description:
Symptoms:
- Monday 27/Apr/2015, 3 out of 10 internal floating IPs were not accessible.
- Amar observed all bad floating IPs were all on one node.
- He then created a new floating IP on that node A, and one on another node B.
- He found that floating IP on node A did not work. However, floating IP on node B worked.
- He has a plausible hypothesis that the vrouter on that node A is messed up. Recommended to restart vrouter service.
- After vrouter restart, the IPs were accessible again.
The node was cp-13.
To manage notifications about this bug go to:
https://bugs.launchpad.net/jio/+bug/1449422/+subscriptions