Open TimJones opened 1 year ago
The Kubernetes project currently lacks enough contributors to adequately respond to all issues.
This bot triages un-triaged issues according to the following rules:
lifecycle/stale
is appliedlifecycle/stale
was applied, lifecycle/rotten
is appliedlifecycle/rotten
was applied, the issue is closedYou can:
/remove-lifecycle stale
/close
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle stale
The Kubernetes project currently lacks enough active contributors to adequately respond to all issues.
This bot triages un-triaged issues according to the following rules:
lifecycle/stale
is appliedlifecycle/stale
was applied, lifecycle/rotten
is appliedlifecycle/rotten
was applied, the issue is closedYou can:
/remove-lifecycle rotten
/close
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle rotten
The Kubernetes project currently lacks enough active contributors to adequately respond to all issues and PRs.
This bot triages issues according to the following rules:
lifecycle/stale
is appliedlifecycle/stale
was applied, lifecycle/rotten
is appliedlifecycle/rotten
was applied, the issue is closedYou can:
/reopen
/remove-lifecycle rotten
Please send feedback to sig-contributor-experience at kubernetes/community.
/close not-planned
@k8s-triage-robot: Closing this issue, marking it as "Not Planned".
/reopen
/remove-lifecycle rotten
@cprivitere: Reopened this issue.
/lifecycle frozen
/triage accepted
/remove-lifecycle frozen
@TimJones Hey did you get an error message from the API when this occurred? Was CPEM continuing to do bad things that caused it to get quickly back into the error state?
Also, does CPEM cause all the BGP to fail by asking for 11 prefixes? I'm trying to understand the value of CPEM throwing the error instead of just passing along the API error.
If we add that setting, we're now making a customer change TWO configs (one via support and one in CPEM) so I really want to understand the value before we do that.
Apologies but I don't have any of the logs from CPEM for incident any more, only from the MetalLB side, which was:
{"caller":"level.go:63","error":"read OPEN from \"169.254.255.1:179\": got BGP notification code 0x0601 (Maximum Number of Prefixes Reached)","level":"error","localASN":65000,"msg":"failed to connect to peer","op":"connect","peer":"169.254.255.1:179","peerASN":65530,"ts":"2023-09-08T14:10:04.171904229Z"}
We recently ran into an issue in one of our clusters in that we had more than 10 LoadBalancer Services which tried to peer via MetalLB BGP. Due to a default limit of 10 prefixes, this caused all subsequent BGP peering to fail until we had Equinix support reset the BGP session after removing the excess Services (switched to using shared IPs).
Since this is a limit in Equinix side, I think it would be of value for the Equinix CCM to also limit number of IPs/prefixes that can be requested/ordered. Ideally an arg that can be set if the customer has arranged a higher session limit.
See https://deploy.equinix.com/developers/docs/metal/bgp/bgp-on-equinix-metal/#bgp-prefix-limit