Last Modified: Jun 17, 2020
See more info
Known Affected Versions:
14.0.0, 13.1.0, 13.0.1, 13.0.0, 12.1.3, 12.1.2
Opened: Jun 08, 2018
Using CGNAT or FW-NAT on a cluster may cause a TMM core if there are intra-cluster communication issues that cause CMP state transitions. The system writes messages to /var/log/tmm* similar to the following: notice CDP: exceeded 1/2 timeout for PG 1 notice CDP: PG 1 timed out notice CDP: New pending state 0f -> 0d notice Immediately transitioning dissaggregator to state 0xd notice cmp state: 0xd notice CDP: New pending state 0d -> 0f ... notice cmp state: 0xf notice CDP: exceeded 1/2 timeout for PG 1
Traffic disrupted while tmm restarts.
-- A LSN pool or FW-NAT source translation that has persistence enabled. -- Intra-cluster communication issues that cause CMP state transitions.
There is no workaround at this time.
TMM does not crash when an unexpected remote session DB response is received.