Cisco Bug: CSCum58876 - RPF change is not updated correct or mfdm/ipfib crash if oifl > 40 svis
May 20, 2019
- Cisco Nexus 9000 Series Switches
Known Affected Releases
<B>Symptom:</B> The QA-tested scalability limit for multicast on the Nexus 9500 is 40 outgoing interfaces (OIFs) per multicast route (i.e. per (*,G) or (S,G) entry). When this is exceeded, unexpected behavior has been observed, such as: - RPF update failure - if the RPF path for a multicast source moves from a L3 interface to a VLAN interface (SVI), this is not correctly updated in the FIB and in hardware programming. As a result, streams from this source are punted to the supervisor due to RPF failure. - The 'mfdm' service may crash on the supervisor, causing a HAP reset of one or both supervisors. - The 'ipfib' service may crash on any or all line cards and/or fabric modules, causing a HAP reset of the affected module(s). - MTS exhaustion may be observed. <B>Conditions:</B> The noted issues are known to occur when the OIF list (OIFL) exceeds 40 entries for one or more multicast routes. The device is not guaranteed to experience these crashes or programming failures once in this state. However, once in this state, multicast churn (any actions that would require reprogramming of multiple OIF lists in hardware) can cause the aforementioned issues to appear.
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