Bug ID 425568: MySQL monitor may hang under repeated connection failures

Last Modified: Oct 10, 2018

Bug Tracker

Affected Product:  See more info
BIG-IP GTM, LTM(all modules)

Known Affected Versions:
10.2.1, 11.1.0, 11.2.0, 11.2.1, 11.3.0, 11.4.0, 11.4.1

Fixed In:
11.5.0, 11.4.1 HF9, 11.3.0 HF9

Opened: Jul 18, 2013
Severity: 3-Major
Related AskF5 Article:
K15242

Symptoms

After receiving errors from a SQL server, which causes the server to close the connection, the monitor (client) tries to continue to use the cached database connection that is no longer valid.

Impact

Monitor marks pool member down when it may not be down.

Conditions

Under certain rare conditions, when a database monitor is unable to connect repeatedly (locked out for too many failed passwords, for example), the database monitor may get into a state where it becomes unresponsive and stop trying to send monitor probes.

Workaround

To work around this issue, you can change the count setting of the affected MySQL or MSSQL monitor to a non-zero value that specifies the number of monitor probes to be sent per connection to the database. A count value of zero (0), which is the default value, indicates the BIG-IP system will leave a single connection open to send probes.

Fix Information

This release contains a fix for the intermittent hang of SQL monitors when cached database connections are closed on the server.

Behavior Change