上午收到告警，一台 dataguard 宕机了。软件版本为 11.2.0.4. 登录上来首先查看 alert 文件 .
| Fri Jan 25 11:00:39 2019 System State dumped to trace file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_ora_42298.trc Fri Jan 25 11:00:54 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_ora_42298.trc: Fri Jan 25 11:00:58 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:01:19 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:01:39 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:01:59 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:02:19 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:02:40 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:03:00 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:03:21 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:03:41 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:04:01 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:04:21 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:04:41 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:05:01 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:05:21 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:05:27 2019 RFS[3]: Selected log 9 for thread 1 sequence 85314 dbid -1036715693 branch 899829907 Fri Jan 25 11:05:27 2019 Media Recovery Waiting for thread 1 sequence 85314 (in transit) Recovery of Online Redo Log: Thread 1 Group 9 Seq 85314 Reading mem 0 Mem# 0: /san/data/oradata/EIN1/standby_redo03.log Fri Jan 25 11:05:27 2019 Archived Log entry 165577 added for thread 1 sequence 85313 ID 0xc2352453 dest 1: Fri Jan 25 11:05:41 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:06:01 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:06:21 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:06:41 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_pmon_56104.trc: ORA-03170: deadlocked on readable physical standby (undo segment 65535) Fri Jan 25 11:06:45 2019 Errors in file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_lgwr_56177.trc: ORA-04020: deadlock detected while trying to lock object SYS.ERAC LGWR (ospid: 56177): terminating the instance due to error 4020 Fri Jan 25 11:06:46 2019 System state dump requested by (instance=1, osid=56177 (LGWR)), summary=[abnormal instance termination]. System State dumped to trace file /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_diag_56159_20190125110646.trc
|
可以看到持续大量的 ORA-03170 死锁告警， 11:06:46 分的时候实例被终止。在贴出的 alert 日志开头生成了一个 trace 文件，首先分析一下这个文件。
Vi 编辑 /u01/app/oracle/diag/rdbms/eracpdg01/ERAC/trace/ERAC_ora_42298.trc 文件，搜索关键字 final ，定位到下面这段：
| SO: 0xa1cccae30, type: 4, owner: 0xa0c7c8da8 , flag: INIT/-/-/0x00 if: 0x3 c: 0x3 proc=0xa0c7c8da8, name=session, file=ksu.h LINE:12729, pg=0 (session) sid: 2228 ser: 9 trans: (nil), creator: 0xa0c7c8da8 flags: (0x45) USR/- flags_idl: (0x1) BSY/-/-/-/-/- flags2: (0x40009) -/-/INC DID: , short-term DID: txn branch: (nil) edition#: 100 oct: 3, prv: 0, sql: 0xa33cf5e08, psql: 0xa3e7cdec0, user: 627/CONVERT_R ksuxds FALSE at location: 0 service name: SYS$USERS client details: O/S info: user: EM-H6VQXD2$, term: unknown, ospid: 1234 machine: EM-H6VQXD2 program: JDBC Thin Client application name: JDBC Thin Client, hash value=2546894660 Current Wait Stack: 0: waiting for 'library cache lock' handle address=0xa3ff9eb18, lock address=0x50de690a0, 100*mode+namespace=0x1004a0002 wait_id=945878 seq_num=28394 snap_id=1 wait times: snap=2 min 53 sec, exc=2 min 53 sec, total=2 min 53 sec wait times: max=15 min 0 sec, heur=2 min 53 sec wait counts: calls=60 os=60 in_wait=1 iflags=0x15a2 There is at least one session blocking this session. Dumping 1 direct blocker(s): inst: 1, sid: 970, ser: 1 Dumping final blocker: inst: 1, sid: 2059, ser: 26985 Wait State: fixed_waits=0 flags=0x22 boundary=(nil)/-1 Session Wait History: elapsed time of 0.000134 sec since current wait 0: waited for 'SQL*Net message from client' driver id=0x54435000, #bytes=0x1, =0x0 wait_id=945877 seq_num=28393 snap_id=1 wait times: snap=3.470941 sec, exc=3.470941 sec, total=3.470941 sec wait times: max=infinite wait counts: calls=0 os=0 occurred after 0.000001 sec of elapsed time 1: waited for 'SQL*Net message to client' driver id=0x54435000, #bytes=0x1, =0x0 wait_id=945876 seq_num=28392 snap_id=1 wait times: snap=0.000001 sec, exc=0.000001 sec, total=0.000001 sec wait times: max=infinite wait counts: calls=0 os=0 occurred after 0.000048 sec of elapsed time |
可以看到最源头的 blocker 是 sid: 2059, ser: 26985 这个 session 。
这里的 SO 是 0xa1cccae30, type: 4 说明是会话信息 , owner: 0xa0c7c8da8 是这个 SO （ 0xa1cccae30 ）的父节点。
搜索这个位置所在的 PROCESS 进程，发现正是 LGWR 进程
PROCESS 17: LGWR
然后直接搜索 sid: 2059 ser: 26985
| SO: 0xa18befcd0, type: 4, owner: 0xa18786ed8, flag: INIT/-/-/0x00 if: 0x3 c: 0x3 proc=0xa18786ed8, name=session, file=ksu.h LINE:12729, pg=0 (session) sid: 2059 ser: 26985 trans: 0xa1b361290, creator: 0xa18786ed8 flags: (0x41) USR/- flags_idl: (0x1) BSY/-/-/-/-/- flags2: (0x40009) -/-/INC DID: , short-term DID: txn branch: 0xa0f20f998 edition#: 100 oct: 3, prv: 0, sql: 0xa345bb8d0, psql: 0xa2fe3c0a8, user: 510/READER ksuxds FALSE at location: 0 service name: SYS$USERS client details: O/S info: user: Administrator, term: ADMIN-PC, ospid: 8960:8832 machine: WORKGROUP\ADMIN-PC program: plsqldev.exe application name: PL/SQL Developer, hash value=1190136663 action name: SQL 窗口 - ISIN 香港净值 .sql , hash value=506963986 Current Wait Stack: 0: waiting for 'SQL*Net message from dblink ' driver id=0x28444553, #bytes=0x1, =0x0 wait_id=474 seq_num=475 snap_id=1 wait times: snap=4 min 45 sec, exc=4 min 45 sec, total=4 min 45 sec wait times: max=infinite, heur=4 min 45 sec wait counts: calls=0 os=0 in_wait=1 iflags=0x5a0 There are 28 sessions blocked by this session. |
可以看到是 reader 用户从一台 Windows 的 PL/SQL developer 发起的一段 sql 文本导致的。而且这段 sql 使用了 db_link 。已经持续了 4 分 45 秒，接近 5 分钟的时间。该会话阻塞了 28 个会话。
目前为止完整的阻塞情况可能并没有全部得到，但是已经得到了 LGWR 进程被一个 PL/SQL developer 会话所阻塞的结论，在 sql 文本执行了接近五分钟的时候，实例被 LGWR 终止。
这里为什么是五分钟呢？这是因为之前发生过的其他故障，那时修改了一个隐含参数 _adg_parselock_timeout
| SQL> col name for a30 SQL> col value for a20 SQL> col description for a70 SQL> set line 150 SQL> select a.ksppinm name,b.ksppstvl value,a.ksppdesc description 2 from x$ksppi a,x$ksppcv b 3 where a.inst_id = USERENV ('Instance') 4 and b.inst_id = USERENV ('Instance') 5 and a.indx = b.indx 6 and upper(a.ksppinm) LIKE upper('%adg_parselock_timeout%') 7 order by name;
NAME VALUE DESCRIPTION ------------------------------ -------------------- ---------------------------------------------------------------------- _adg_parselock_timeout 30000 timeout for parselock get on ADG in centiseconds _adg_parselock_timeout_sleep 100 sleep duration after a parselock timeout on ADG in milliseconds
SQL> col name for a30; SQL> col value for a10; SQL> select * from 2 (select 3 x.ksppinm name, 4 y.ksppstvl value, 5 y.ksppstdf isdefault, 6 decode(bitand(y.ksppstvf,7),1,'MODIFIED',4,'SYSTEM_MOD','FALSE') ismod, 7 decode(bitand(y.ksppstvf,2),2,'TRUE','FALSE') isadj 8 from 9 sys.x$ksppi x, 10 sys.x$ksppcv y 11 where 12 x.inst_id = userenv('Instance') and 13 y.inst_id = userenv('Instance') and 14 x.indx = y.indx 15 order by 16 translate(x.ksppinm, ' _', ' ')) T where T.name like '%adg_parselock_timeout%';
NAME VALUE ISDEFAULT ISMOD ISADJ ------------------------------ ---------- --------------------------- ------------------------------ --------------- _adg_parselock_timeout 30000 FALSE FALSE FALSE _adg_parselock_timeout_sleep 100 TRUE FALSE FALSE |
_adg_parselock_timeout 该值设置为了 30000 厘秒，约等于 300 秒，也就是五分钟。
搜索 MOS 后确认为 BUG
| Bug 18515268 ORA-4020 in ADG Standby Database causing instance crash by LGWR |
Solution
| Apply interim patch 18515268, if available for your platform and Oracle version. Fixed releases: 11.2.0.4.BP20 12.1.0.2 12.2.0.1 |
来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/31480688/viewspace-2564916/，如需转载，请注明出处，否则将追究法律责任。
转载于:http://blog.itpub.net/31480688/viewspace-2564916/
207
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
