Bug ID 2289817: Erroneous TXQ Full Health Check Marks Interface Down, Causing TMM Pod Connectivity Loss

Last Modified: Oct 27, 2026

Affected Product(s):
BIG-IP NEXT BNK None(all modules)

Known Affected Versions:
2.0.0, 2.0.1, 2.0.2, 2.0.3, 2.1.0, 2.1.1, 2.2.0, 2.2.1, 2.2.2, 2.2.3, 2.2.4, 2.2.5

Opened: Apr 24, 2026

Severity: 3-Major

Symptoms

TMM fails with error, => "Error: Set unhealthy status flag because txq has been full for <####> seconds" => "Error: Marking down interface because unhealthy flag set to true."

Impact

Loss of data-plane connectivity as interfaces are marked down.

Conditions

- TMM handles traffic, but one or more dataplane interfaces (e.g., 1.1, 1.2) expose fewer device queues than the available threads. - In this scenario, multiple TMM threads share a single device queue, which increases the likelihood that the underlying NIC struggles to handle the traffic those threads transmit simultaneously on the TX path. When this happens, one or more device queues temporarily increase their "blocked" counters in the 'tmm/ndal_tx_stats' table. This behavior usually does not cause issues because the NIC almost always recovers on its own. The temporary rise in blocked counters can trigger the internal device TXQ health check, which also uses the number of sent packets to detect when the queue becomes unblocked. If the "sent" counter wraps to zero after it reaches its maximum value, the health check triggers after the next increment of the "blocked" counter.

Workaround

- Disable txq healthcheck by setting the timeout to 0. - In your TMM deployment, add a new configmap or a host file with the following content: ndal tx_dq_timeout_sec 0 - This configmap or host-file should appear within the TMM container as "/opt/lib/tmm/tmm_init.tcl.pre".

Fix Information

None

Behavior Change

Guides & references

K10134038: F5 Bug Tracker Filter Names and Tips