Bug ID 507499: TMM can watchdog under extreme memory pressure.

Last Modified: Nov 07, 2022

Bug Tracker

Affected Product:  See more info
BIG-IP All(all modules)

Known Affected Versions:
11.5.1, 11.5.1 HF1, 11.5.1 HF10, 11.5.1 HF11, 11.5.1 HF2, 11.5.1 HF3, 11.5.1 HF4, 11.5.1 HF5, 11.5.1 HF6, 11.5.1 HF7, 11.5.1 HF8, 11.5.1 HF9, 11.5.10, 11.5.2, 11.5.2 HF1, 11.5.3, 11.5.3 HF1, 11.5.3 HF2, 11.5.4, 11.5.4 HF1

Fixed In:
12.0.0, 11.5.4 HF2

Opened: Feb 17, 2015
Severity: 2-Critical

Symptoms

The TMM can become unresponsive and then be killed by SOD under extreme memory pressure.

Impact

The TMM is restarted; flows that can't failover to a backup node are disrupted. If the killed TMM was not the source of the memory pressure, there may not be enough memory for a new TMM instance to come up.

Conditions

Under extremely high memory pressure, linux will page out anything that isn't nailed down, including the shared memory containing the system-wide logging configuration. When this happens, and something in the TMM considers logging, the TMM will be de-scheduled while the linux kernel tries to swap something else out and swap the configuration page back in. Under such conditions, several seconds may go by before the memory can be swapped back in. SOD detects that the TMM is unresponsive and restarts the TMM.

Workaround

A release that locks the logging configuration into RAM is required to correct the poor response to being out of memory. Note: this change improves system handling in out-of-memory conditions -- it does NOT address any of the potential sources of the out-of-memory condition.

Fix Information

The logging configuration is now locked into RAM.

Behavior Change