[Bug 2155119] Re: Backport upstream patch to avoid 100% CPU on full-routing tables (OVS Issue #185)

[Seyeong Kim]

*** This bug is a duplicate of bug 2109823 ***
    https://bugs.launchpad.net/bugs/2109823

** This bug has been marked a duplicate of bug 2109823
   [SRU] openvswitch 3.3.4 (noble) 3.5.1 (plucky/questing) point release

-- 
You received this bug notification because you are a member of Ubuntu
OpenStack, which is subscribed to openvswitch in Ubuntu.
https://bugs.launchpad.net/bugs/2155119

Title:
  Backport upstream patch to avoid 100% CPU on full-routing tables (OVS
  Issue #185)

Status in openvswitch package in Ubuntu:
  New

Bug description:
  Affects Package: openvswitch (Jammy Cloud Archive - Caracal)
  Current Version: 3.3.0-1ubuntu3.1~cloud0

  Description:
  When running a BGP daemon (like Bird or FRR) handling a full routing table on the same host and namespace as Open vSwitch, ovs-vswitchd spins at 100% CPU due to continuous Netlink route-table cache invalidations from non-standard routing tables.

  This was fixed upstream in the OVS 3.3 branch starting with version
  3.3.1, but the current Caracal UCA package is stuck on a patched 3.3.0
  core and is missing this critical performance fix.

  Upstream References:

  GitHub Issue: https://github.com/openvswitch/ovs-issues/issues/185

  Upstream Commit: "route-table: Avoid routes from non-standard routing
  tables." (Committed March 2024, native to 3.3.1+)

  Request:
  Please cherry-pick/backport this specific patch into the Ubuntu Jammy Caracal package update stream to prevent control plane CPU exhaustion.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/openvswitch/+bug/2155119/+subscriptions