(FAQ)DB2 drop table非常慢
原因分析： 1.通过db2 update monitor switches using bufferpool on sort on lock on statement on table on uow on . db2 get snapshot for locks on DB_NAME 结果发现在系统表上有锁: Application handle = 115 l Application ID = l Sequence number = 0001 l Application name = l CONNECT Authorization ID = l Application status = Lock-wait l Status change time = l Application code page = 1386 l Locks held = 22 l Total wait time (ms) = 70945 l l ID of agent holding lock = 114 l Application ID holding lock = *LOCAL.DB2.070629111515 l Lock name = 0x00000500FF0100000000000052 l Lock attributes = 0x00000010 l Release flags = 0x40000000 l Lock object type = End of Table l Lock mode = Update Lock (U) l Lock mode requested = Update Lock (U) l Name of tablespace holding lock = SYSCATSPACE l Schema of table holding lock = SYSIBM l Name of table holding lock = SYSCOLPROPERTIES l Lock wait start timestamp = 2007-06-29 19:15:53.193235 l 2.通过改变maxlocks和locklist等都没有改善. 3.通过分析、查看，发现在32个节点中,每个NODExxxx/SQL00001中的db2rhist.asc和db2rhist.bak都达到300MB.应该是时间花在这里 解决方法： 1.把db2rhist.asc和db2rhist.bak删除. 2.update db cfg using REC_HIS_RETENTN nn nn为您需要保留的天数 PRUNE HISTORY timestamp [WITH FORCE OPTION]