After primary log archiving method fails I expected my logs to archive to the FAILARCHPATH but this failed as well. Why?
DB2 will try to write (archive logs) to the FAILARCHPATH whenever there is a media error with the primary log archiving method (LOGARCHMETH1). DB2 will retry archiving to LOGARCHMETH1 a certain number of times based on what you have set in NUMARCHRETRY. After this is exhausted, DB2 will then attempt to archive to the local path specified in FAILARCHPATH (as long as this is set). If this fails you may see a message similar to:
2009-08-08-08.12.45.111823-300 E54716A491 LEVEL: Warning
PID : 19999 TID : 12 PROC : db2sysc 0
INSTANCE: db2inst1 NODE : 000
EDUID : 15 EDUNAME: db2logmgr (RTSDC) 0
FUNCTION: DB2 UDB, data protection services, sqlpgArchiveLogFile,
MESSAGE : ADM1848W Failed to archive log file "S0000004.LOG" to
When you set FAILARCHPATH for the first time the full path will be created. If, for whatever reason, a directory in the path gets removed at some point, DB2 will recreate the full directory structure again when it comes time to archive logs to the FAILARCHPATH. What this means is that you can see the above error when the permissions or the owner are mistakenly changed, causing any "writes" by the instance owner to fail.