How to avoid frequent log table switching with the Symantec Endpoint Protection Manager (SEPM)?
In the scm-server-0.log / scm-server-1.log log files of the SEPM, frequent lines such as these can be observed:
2012-04-02 15:34:45.063 THREAD 27 WARNING: Log table switched to: SERVER_CLIENT_LOG_1, old table estimated row count: 4055, to add row count in new table: 2, last switch time: 2012-04-02 15:16:08
2012-04-02 15:35:10.441 THREAD 39 WARNING: Log table switched to: AGENT_TRAFFIC_LOG_1, old table estimated row count: 20751, to add row count in new table: 46, last switch time: 2012-04-02 15:22:08
2012-04-02 15:49:23.570 THREAD 39 WARNING: Log table switched to: AGENT_TRAFFIC_LOG_2, old table estimated row count: 1181, to add row count in new table: 18, last switch time: 2012-04-02 15:35:05
2012-04-02 16:10:24.138 THREAD 39 WARNING: Log table switched to: AGENT_TRAFFIC_LOG_1, old table estimated row count: 5081, to add row count in new table: 100, last switch time: 2012-04-02 15:49:23
The lines may mention SEP client log tables (AGENT_TRAFFIC_LOG_1/2, AGENT_PACKET_LOG_1/2, AGENT_BEHAVIOR_LOG_1/2, AGENT_SECURITY_LOG_1/2, AGENT_SYSTEM_LOG_1/2), SEPM client log tables (SERVER_CLIENT_LOG_1/2) or possibly Enforcer log tables (ENFORCER_CLIENT_LOG_1/2, ENFORCER_TRAFFIC_LOG_1/2).
For log storage tables in the database the SEPM alternates between an *_1 and an *_2 table. When the _1 table is full the SEPM will truncate the _2 table and continue writing to it until it too is full, then once again truncate and switch to using the _1 table as "active". For further information please see Manage log data in Endpoint Protection Manager.
To avoid using resources on the SEPM server as well as the database server for uploading and storing large amounts of log entries that will soon be overwritten, there are three options: