For more information regarding the security incident at F5, the actions we are taking to address it, and our ongoing efforts to protect our customers, click here.

Forum Discussion

mr_shaggy_17493's avatar
mr_shaggy_17493
Icon for Nimbostratus rankNimbostratus
Aug 24, 2016

High Memory Used on TMM Instance

Hi All,

 

I have a pair of BIG-IP 5200v. Today I just realise that on the last 30 days, TMM memory utilization on my F5 active unit is getting high, and keep increase up to 50% used. Normally, its only 4-5% of TMM memory used. Anyone ever face the same issue? Below is the capture:

 

admin@(f5-1)(cfg-sync In Sync)(Active)(/Common)(tmos) show sys tmm-info

 

-- Sys::TMM: 0.0 --

 

Global TMM Process Id 12664 Running TMM Id 0 TMM Count 1 CPU Id 0

 

Memory (bytes) Total 13.4G Used 7.0G

 

CPU Usage Ratio (%) Last 5 Seconds 3 Last 1 Minute 3 Last 5 Minutes 3

 

-- Sys::TMM: 0.2 --

 

Global TMM Process Id 12664 Running TMM Id 2 TMM Count 1 CPU Id 2

 

Memory (bytes) Total 0 Used 0

 

CPU Usage Ratio (%) Last 5 Seconds 1 Last 1 Minute 1 Last 5 Minutes 1

 

-- Sys::TMM: 0.4 --

 

Global TMM Process Id 12664 Running TMM Id 4 TMM Count 1 CPU Id 4

 

Memory (bytes) Total 0 Used 0

 

CPU Usage Ratio (%) Last 5 Seconds 1 Last 1 Minute 1 Last 5 Minutes 1

 

-- Sys::TMM: 0.6 --

 

Global TMM Process Id 12664 Running TMM Id 6 TMM Count 1 CPU Id 6

 

Memory (bytes) Total 0 Used 0

 

CPU Usage Ratio (%) Last 5 Seconds 2 Last 1 Minute 2 Last 5 Minutes

 

As seen on the capture above, TMM instance 0.0 using 7.0G of memory from the 13.4G allocated. Is there a way to know what process running under TMM instance 0.0? I want to know what process that was using memory so high, or whether it was a memory leak, etc.

 

Thanks anyway,

 

Ahmad

 

5 Replies

  • JG's avatar
    JG
    Icon for Cumulonimbus rankCumulonimbus

    I am experiencing a very similar issue myself, and have an open case with F5 Support. I'd suggest that you do the same, with a qkview taken during such an event. And which BIG-IP version are you running?

     

  • There were memory leak issues on 11.5.4 + HF1 that was fixed in the 11.5.4 + HF2 code version. There is a chance 11.6 code train could be suffering from the same issue that may be fixed in the latest HF for 11.6

     

    • matt_64003's avatar
      matt_64003
      Icon for Cirrus rankCirrus

      Did F5 make any progress with your case?

       

      We're running 11.6.0 HF6 and it looks like TMM has a memory leak. After about two months it reaches 80% and our apps start slowing down and timing out. A reboot fixes it, for another two months. I have been unable to get a qkview before someone reboots though so I don't have much info for F5 support.

       

      Thanks.

       

    • mr_shaggy_17493's avatar
      mr_shaggy_17493
      Icon for Nimbostratus rankNimbostratus

      Hi matt,

       

      I think they didn't found the exact root cause for the issue and just suggest me to switch boot to another partition. This solution is worked for me, and memory statistics are getting normal after the switch boot. But now it's HA pair are looking to start the same symptom here, where the memory statistics are gradually higher (previously it's only happen on one unit in the HA pair).

       

      As the BIG-IP 11.6.0 has reached it's EoSD, so I plan to upgrade it into the higher version. We are using ASM on this unit, so besides of this issue, we also need to periodically update the attack signature, and there will be no more update if we keep using this OS version.

       

      Thanks,

       

      Ahmad