Cisco Bug: CSCvi12661 - BGP Crash or stale info on standby RSP with aggregate config and aggregate prefix learned from peer.
May 25, 2020
- Cisco ASR 9000 Series Aggregation Services Routers
Known Affected Releases
Symptom: 1.BGP Crash on standby RSP or 2. Stale routing information on standby RSP for aggregate addresses (V4/V6) Way to detect is, to compare for # of available paths for aggregate-addresses in active vs standby RSP sh bgp [vrf ID] ipv4 uni IPv4_PFX | i avail and sh bgp [vrf ID] ipv4 uni IPv4_PFX stand | i avail OR sh bgp [vrf ID] ipv6 uni IPv6_PFX | i avail and sh bgp [vrf ID] ipv6 uni IPv6_PFX stand | i avail look for paths available (how many) and best path index. It needs to match b/w Active and Stby. RP/0/RP0/CPU0:#sh bgp vrf <> ipv4 uni <> | i avail Paths: (17 available, best #1) RP/0/RP0/CPU0:#sh bgp vrf <> ipv4 uni <> stand | i avail Paths: (17 available, best #1) RP/0/RP0/CPU0:#sh bgp ipv4 uni <> | i avail Paths: (3 available, best #3) RP/0/RP0/CPU0:#sh bgp ipv4 uni <> stand | i avail Paths: (8 available, best #1) Conditions: The issue is seen when we have an aggregate route is configured and following event occurs: Initially prefix has 2 paths: Path1: aggregate Path2: learned from neighbor Path2 gets withdrawn. Path2 is re-advertised. On standby we add another(redundant) aggregate path via NSR sync code path if standby processes this NSR sync event before network event. So this is based on timing condition. standby bgp crashes if redundant path count wraps unsigned integer value. Note: This issue will not happen if NSR is disabled for bgp process.
Bug details contain sensitive information and therefore require a Cisco.com account to be viewed.
Bug Details Include
- Full Description (including symptoms, conditions and workarounds)
- Known Fixed Releases
- Related Community Discussions
- Number of Related Support Cases