1. 版本描述:
操作系统:AIX 4.3
中间件:CICS4.2 + PTF9
数据库:DB2 UDB v6.1 + PTF9
2. 故障描述
2007-03-11上午出现了2次cics region 异常down的现象.
3. 故障分析
3.1 检查cics region console日志
分析:通过检查cics console日志,也只能发现由于应用程序出现A158错误,并产生了coredump。接着cics region异常down,但并不能知道是那个应用程序引起。二次的Console日志显示如下:
第一次的日志:
ERZ052004I/0602 03/11/07 09:07:36 ywrun : Dump to 'A1580026.dmp' started. ERZ052013I/0409 03/11/07 09:07:37 ywrun : Dump 0026 - 1 core dump(s) created. ERZ052007I/0604 03/11/07 09:07:37 ywrun : Dump to 'A1580026.dmp' completed. ERZ014016E/0036 03/11/07 09:07:37 ywrun : Transaction 'CPMI', Abend 'A158', at '????'. ERZ057002E/0140 03/11/07 09:07:37 ywrun : CICS internal error: Unsuccessful condition '(tran_TidEqual(ThreadTid, OldTid))' for function 'TasLU_ISyncpoint' on line 968 in file '/cics/PTF9/src/tas/lu/src/taslusyn.c' of module 16 ERZ057005E/0140 03/11/07 09:07:37 ywrun : CICS internal error: Abnormally terminating region with exit code 'U5701' ERZ010003I/0094 03/11/07 09:07:37 ywrun : CICS is performing region abnormal termination in process 'cicsas' ERZ052004I/0602 03/11/07 09:07:37 ywrun : Dump to 'SYSA0027.dmp' started. ERZ052007I/0604 03/11/07 09:07:41 ywrun : Dump to 'SYSA0027.dmp' completed. ERZ010089E/0367 03/11/07 09:07:41 ywrun : Application Server 108 ended unexpectedly ERZ010144I/0375 03/11/07 09:07:41 ywrun : Application server 104 started ERZ010144I/0375 03/11/07 09:07:41 ywrun : Application server 105 started ERZ014016E/0036 03/11/07 09:07:41 ywrun : Transaction 'CPMI', Abend 'A147', at '????'. |
第二的日志:
ERZ052004I/0602 03/11/07 10:13:02 ywrun : Dump to 'A1580001.dmp' started. ERZ052013I/0409 03/11/07 10:13:02 ywrun : Dump 0001 - 1 core dump(s) created. ERZ052007I/0604 03/11/07 10:13:02 ywrun : Dump to 'A1580001.dmp' completed. ERZ014016E/0036 03/11/07 10:13:02 ywrun : Transaction 'CPMI', Abend 'A158', at '????'. ERZ057002E/0140 03/11/07 10:13:02 ywrun : CICS internal error: Unsuccessful condition '(tran_TidEqual(ThreadTid, OldTid))' for function 'TasLU_ISyncpoint' on line 968 in file '/cics/PTF9/src/tas/lu/src/taslusyn.c' of module 16 ERZ057005E/0140 03/11/07 10:13:02 ywrun : CICS internal error: Abnormally terminating region with exit code 'U5701' ERZ010003I/0094 03/11/07 10:13:02 ywrun : CICS is performing region abnormal termination in process 'cicsas' ERZ052004I/0602 03/11/07 10:13:02 ywrun : Dump to 'SYSA0002.dmp' started. ERZ052007I/0604 03/11/07 10:13:04 ywrun : Dump to 'SYSA0002.dmp' completed. ERZ010012I/0194 03/11/07 10:13:04 ywrun : CICS is performing an immediate shutdown of region 'ywrun' |
3.2 检查应用日志
分析:通过检查应用日志,2次出现故障的时间点,都只出现了交易7118的错误。从这点看当时出现故障的原因应该和7118交易有关系。应用日志都出现了应用日志显示如下:
第一次日志:
03-11 09:07:36| 492|ckn11021 |1102|000000|12|
03-11 09:07:36| 183|main |1102|000000|机构[11070720]操作员[1047]交易[1102]结束|
03-11 09:07:36|1591|midtelegram|7118|111111|读socket出错,从电信读回包头失败!|
第二次日志:
03-11 10:13:01|1591|midtelegram|7118|111111|读socket出错,从电信读回包头失败!|
3.3 分析A158dump文件
使用cicsdfmt –r ywrun A158??????.dmp01命令格式化产生的A158 dmp文件。从2次dump中都能发现如下信息,所以基本定位在7118交易引起的。需要应用人员进一步分析。
Program Control Information:
Buffer Address = 0xa015ce9c
Signature = ERZ15PCI
Previous Level = 0xa015d8dc
Load Address = 0x20604df0
Current Program:
Use Control = 0xa00375bc
Comm Area = 0x2012b338
Length = 32000
Data Length = 32000
Transid for link =
SyncOnReturn = FALSE
Program Name = md71181
Index into cache pointer array = 744
Current version of the program = 1
Value of Resident attribute = No
Program full path name = /var/cics_regions/ywrun/bin/md71181
4.建议
通过上面的分析,下一步需要应用人员对7118交易的应用程序,进行故障分析。
[@more@]来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/67233/viewspace-904902/,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/67233/viewspace-904902/