Oracle已经要发布10.2.0.3的Patch了, 在Metalink上已经可以找到Bug修复列表了. 这个列表很长, 我初略地看了一下, 抄出以下面的一些我认为需要关注一下的, 从这份列表来看, 10.2.0.2是很不成熟的, Oracle会建议立马打补丁到最新的版本.
3279497 Interupted (Ctrl-C) TRUNCATE can corrupt ASSM table 4188324 Wasted space possible in ASSM segments 4323868 INDEX hints can lead to INDEX SCAN FULL 4378206 Direct load of compressed table can load duplicate rows 5212539 LOB corruption for NOCACHE LOBs (ORA-1555/ORA-22924) 5017007 MVIEW LOB content incorrect after refresh 5458753 SQL may execute in wrong schema. This bug is alerted in Note 392673.1 4963106 SQLLDR SKIP skips wrong number of records 4748597 Moving a LOB in a partitioned table fails to update IND$ 4732503 Self-deadlock on TT enqueue 4699610 CKPT can self deadlock waiting for MTTR state object which it holds 5016142 SMON rollback of large transactions blocks other actions 5068048 SQL*loader hang on direct path load on multi CPU machine 4859002 PMON crash cleaning up serially_reusable PLSQL 4726645 MMON has a slow memory leak ("koh-kghu sessi") 5004782 Memory leak on repeated connect/disconnect 4968994 Poor plan using SYSDATE in predicate over dblink 4461334 Index subpartitions do not inherit index level logging attribute 5448588 Slow LGWR sync performance for large redo write sizes 4590425 RMAN exit status is 0 when validate errors 4994956 Number of configured shared pool subpools not correct 4361163 It is not possible to have > 1000 shared servers on some platforms 4908162 Suboptimal plan possible for partitioned table with many joins 5014836 INSERT as SELECT plan can differ from SELECT plan
我没有看RAC, OLAP等这些不常用的组件的Bug, 只关注了很核心的部分.
留言 (2)
Wasted space possible in ASSM segments
好多版本都有这个问题
看来 Oracle ASSM 的算法还是没搞定
Posted by Fenng | Nov 8, 2006 4:53 PM
我已经上了最新的patch,可是Memory leak问题依然存在,没有被修复,我的Memory不断被oracle.exe吃掉,直到完为止
Posted by kevensun | May 22, 2007 10:06 AM