Skip to content

Conversation

oneswig
Copy link
Member

@oneswig oneswig commented Nov 10, 2022

OVS bridge interfaces drop packets during normal operation, generating lots of spurious alerts. Change the regex to filter out interfaces that don't matter for packet drops.

OVS bridge interfaces drop packets during normal operation.  Change
the regex to filter out interfaces that don't matter for packet
drops.
@oneswig oneswig requested a review from cityofships November 10, 2022 11:07
@oneswig oneswig requested a review from a team as a code owner November 10, 2022 11:07
Copy link
Member

@dougszumski dougszumski left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This should help reduce alert spam, thanks Stig.

# alert on nic packet errors and drops rates > 1 packet/s
- alert: NetworkPacketsDropped
expr: irate(node_network_receive_drop_total{device=~"en.*|eth.*"}[5m]) + irate(node_network_transmit_drop_total{device=~"en.*|eth.*"}[5m]) > 1
expr: irate(node_network_receive_drop_total{device!~"lo|br.*|.*-ovs|tap.*"}[5m]) + irate(node_network_transmit_drop_total{device!~"lo|br.*|.*-ovs|tap.*"}[5m]) > 1
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This will catch, among others:

bond.*
docker0
genev_sys_6081
ovs-system
p-brbond.*
qbr.*
qvb.*
vxlan_sys_4789

The previous regex has already been filtering out OVS bridges - I thought we were only expanding it with |bond.*.

Does by the way |ib.* make sense too?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Perhaps this needs to be addressed. Could be a follow up?

@markgoddard markgoddard merged commit a792e4e into stackhpc/yoga Nov 16, 2022
@markgoddard markgoddard deleted the packet-drops branch November 16, 2022 12:05
@markgoddard
Copy link
Contributor

Backported to Xena: #233

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants