Heim  >  Artikel  >  Datenbank  >  ORA-15041,ORA-01274 故障解决实例

ORA-15041,ORA-01274 故障解决实例

WBOY
WBOYOriginal
2016-06-07 15:53:272298Durchsuche

和客户沟通了下,他们打算把DG停掉,暂时不用了,因此也没有做进一步处理,如果不想让归档日志传递到备库,可以设置主库参数log_

上周五去客户处巡检时,发现原来配置的DG备库未与主库同步,特此记录一下

在5月14日13:58出现故障,3398日志无法应用,以下是alert日志

Thu May 14 13:58:01 2015
Errors in file /u01/app/Oracle/diag/rdbms/tcdg/tc/trace/tc_pr00_3473760.trc:
ORA-01119: error in creating database file '+data'
ORA-17502: ksfdcre:4 Failed to create file +data
ORA-15041: diskgroup "DATA" space exhausted
File #108 added to control file as 'UNNAMED00108'.
Originally created as:
'+DATA/tc/datafile/oa2015.387.879688649'
Recovery was unable to create the file as:
'+data'
MRP0: Background Media Recovery terminated with error 1274
Errors in file /u01/app/oracle/diag/rdbms/tcdg/tc/trace/tc_pr00_3473760.trc:
ORA-01274: cannot add datafile '+DATA/tc/datafile/oa2015.387.879688649' - file could not be created
Managed Standby Recovery not using Real Time Apply
Recovery interrupted!
Recovered data files to a consistent state at change 10293976250522
Thu May 14 13:58:06 2015
MRP0: Background Media Recovery process shutdown (tc)

由于空间不足,出现了ORA-15041错误,然后MRP0进程被终止

再来看trace文件内容:

*** 2015-05-14 13:58:01.562
----- START Event Driven Actions Dump ----
---- END Event Driven Actions Dump ----
----- START DDE Actions Dump -----
Executing SYNC actions
Executing ASYNC actions
----- END DDE Actions Dump (total 0 csec) -----
ORA-01119: error in creating database file '+data'
ORA-17502: ksfdcre:4 Failed to create file +data
ORA-15041: diskgroup "DATA" space exhausted
File #108 added to control file as 'UNNAMED00108'.
Originally created as:
'+DATA/tc/datafile/oa2015.387.879688649'
Recovery was unable to create the file as:
'+data'
*** 2015-05-14 13:58:01.598 4329 krsh.c
MRP0: Background Media Recovery terminated with error 1274
ORA-01274: cannot add datafile '+DATA/tc/datafile/oa2015.387.879688649' - file could not be created
*** 2015-05-14 13:58:01.613 4329 krsh.c
Managed Standby Recovery not using Real Time Apply

*** 2015-05-14 13:58:01.616
MRP: Prodding archiver at standby for thread 1 seq 3398
----- Redo read statistics for thread 1 -----
Read rate (ASYNC): 1631875714Kb in 8455733.63s => 0.19 Mb/sec
Total redo bytes: 1631903362Kb Longest record: 53Kb, moves: 546631/3909060988 moved: 2869Mb (0%)
Longest LWN: 61559Kb, reads: 53311640
Last redo scn: 0x095c.c0cc389a (10293976250522)
Change vector header moves = 505014709/3028421495 (1%)
----------------------------------------------

*** 2015-05-14 13:58:01.626
Media Recovery drop redo thread 1
KCBR: Redo cache copies/changes = 7576594/7576586
Wait to push change maps to slaves = 136500s

*** 2015-05-14 13:58:02.545
Completed Media Recovery
Checking to start in-flux buffer recovery from SCN 2396.3234383719 to SCN  (non-inclusive) 2396.3234609306
Influx recovery found in-flux buffers

*** 2015-05-14 13:58:02.622
Influx Media Recovery add redo thread 1
Managed Standby Recovery: Standby online log for thr 1  seq 3397 not found. Looking whether archived..
Looking for archived log thr 1 seq 3397 recovery branch id 869789191 that contains SCN 10293976024935
Checking whether scn 10293976024935 in los 10293976024935 and nxs 10293976125751
Managed Standby Recovery: Opening archived log /oraarch/ARC_CRM5_3397_869789191_1.log during invocation of recoverable recovery

*** 2015-05-14 13:58:03.937
Resized overflow buffer to 2435K (for 2435K LWN)
Resized overflow buffer to 3923K (for 3923K LWN)
Resized overflow buffer to 14242K (for 14242K LWN)

*** 2015-05-14 13:58:06.105
MRP: Prodding archiver at standby for thread 1 seq 3398
Managed Recovery: Not Active posted.

由于redo apply的进程被终止,因此应用到3397归档后,之后的日志都未被应用,但是归档日志都已经顺利传到了备库

SQL> select thread#,sequence#,applied from v$archived_log where sequence#>3391;

  THREAD#  SEQUENCE# APPLIED
---------- ---------- ---------
        1      3392 YES
        1      3393 YES
        1      3394 YES
        1      3395 YES
        1      3396 YES
        1      3397 YES
        1      3398 NO
        1      3399 NO
        1      3400 NO
        1      3401 NO
        1      3402 NO
        1      3403 NO
        1      3404 NO
        1      3405 NO
        1      3406 NO
        1      3407 NO
        1      3408 NO
        1      3409 NO
        1      3410 NO
        1      3411 NO
        1      3412 NO
        1      3413 NO

还可以通过下面的SQL语句去分析备库落后了多少归档日志未完成应用。

Stellungnahme:
Der Inhalt dieses Artikels wird freiwillig von Internetnutzern beigesteuert und das Urheberrecht liegt beim ursprünglichen Autor. Diese Website übernimmt keine entsprechende rechtliche Verantwortung. Wenn Sie Inhalte finden, bei denen der Verdacht eines Plagiats oder einer Rechtsverletzung besteht, wenden Sie sich bitte an admin@php.cn