Forum Discussion

Satriaji's avatar
Satriaji
Icon for Cirrus rankCirrus
Mar 23, 2023

TMM Restarting (OS Version 12.1.1)

Hi Everyone, I have a problem with my F5 device. The problem is the tmm service always restarting continuesly, so its make the device INOPERATIVE (intermittent active-standby).  Here the logs : "Ma...
application delivery
inopperative
tmm
tmm restart
CA_Valli's avatar
CA_Valli
Icon for MVP rankMVP

If the unit is covered by a support contract and this is service impacting, your best bet is to open a support case to investigate the issue live.

Most common reason for a unit to behave like this is if you have a VLAN failover enabled, and unit has the monitored links down.
I haven't seen this specific evidence in your logs, but you do have a deamon heartbeat failure error. That means a system process is failing to respond to HA daemon heartbeat messages, which triggers the failover. 

Again, I'd suggest you have support look into this. 

Satriaji's avatar
Satriaji
Icon for Cirrus rankCirrus
Mar 23, 2023

The technical support for this OS device has been out of support (Technical EOS at 2021), so we cannot open TAC to support.

For now the log still appear :

[msi@JTLSF-DNS-PR-SOA-PREPROD:INOPERATIVE:Standalone] ~ # Mar 23 23:51:45 JTLSF-DNS-PR-SOA-PREPROD emerg logger: Re-starting tmm

 

This device is vCMP Guest, the other guest is normal. Just thist guest that having problem tmm restarting.

 

 

 

  • DuncanJones's avatar
    DuncanJones
    Icon for Nimbostratus rankNimbostratus
    Mar 24, 2023
    Satriaji wrote:

    The technical support for this OS device has been out of support (Technical EOS at 2021), so we cannot open TAC to support.

    For now the log still appear : Tower Defense

    [msi@JTLSF-DNS-PR-SOA-PREPROD:INOPERATIVE:Standalone] ~ # Mar 23 23:51:45 JTLSF-DNS-PR-SOA-PREPROD emerg logger: Re-starting tmm

     

    This device is vCMP Guest, the other guest is normal. Just thist guest that having problem tmm restarting.

     

     

     

    The Traffic Management Microkernel (tmm) process may restart without a stack trace or core file after becoming disconnected from the Master Control Program Daemon (mcpd) process.

    This issue occurs when all of the following conditions are met:

    • The Linux kernel memory becomes fragmented.
    • Another memory-intensive operation occurs, which may include:
      • You perform a configuration synchronization (ConfigSync) with full reload.
      • You run a tcpdump session.

    If tmm fails to start or restart properly as a result of this issue, you can try to recover the system temporarily by restarting the BIG-IP system.