Bug ID 561976: Values of high-water and low-water mark for 'apd' pending request queue might not handle requests completely.

Last Modified: Apr 10, 2019

Bug Tracker

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

Known Affected Versions:
11.6.0, 11.6.0 HF1, 11.6.0 HF2, 11.6.0 HF3, 11.6.0 HF4, 11.6.0 HF5, 11.6.0 HF6, 11.6.0 HF7, 11.6.0 HF8

Fixed In:
11.6.1

Opened: Dec 08, 2015
Severity: 3-Major
Related AskF5 Article:
K55561335

Symptoms

Under heavy authentication requests from tmm with a slow or down back-end authentication server, the apd accept connection queue could get full, resulting in apd logs: AD module: authentication with '1439805563539620' failed: Too many open files.

Impact

Authentication failures; might bring authentication rate down to zero.

Conditions

- Incoming authentication request to apd (from tmm) is very high. - Back-end authentication server is slow or down.

Workaround

Adjust the value of connhwm, connlwm and soconnmax values using tmsh commands. - To set the value to 1024, use the following command: sysctl -w net.core.somaxconn=1024. - Change Low water mark first using the following command: tmsh modify sys db apm.apd.connlwm value 480. - Change highwater mark next using the following command: tmsh modify sys db apm.apd.connhwm value 512.

Fix Information

Values of high-water and low-water mark for the 'apd' pending request queue now handle requests as expected.

Behavior Change