informix数据库故障报逻辑日志回滚失败，急，在线等！

lansworld

今天数据库因为磁盘阵列故障导致informix数据库不正常关闭，再次启动后(oninit -pv)数据库报错，检查错误日志如下

19:27:30  Physical Recovery Started at Page (2:182885).
19:28:00  Physical Recovery Complete: 86106 Pages Examined, 86099 Pages Restored.
19:28:01  Logical Recovery Started.
19:28:01  10 recovery worker threads will be started.
19:32:55  Logical Recovery has reached the transaction cleanup phase.
19:32:59  Assert Failed: Rollback error 126
19:32:59  IBM Informix Dynamic Server Version 9.40.FC7
19:32:59   Who: Session(16, informix@zjnpmdb, 0, 5053c887
                Thread(97, fast_rec, 505386da0, 1)
                File: rstrans.c Line: 2437
19:32:59   Results: Log record (OLDRSAM:HDELETE) in log 221077, offset 0x52701b4 was not rolled back
19:32:59   Action: Use 'onlog' to view the transaction and repair manually.
19:32:59  stack trace for pid 23350 written to /tmp/af.449336a
19:32:59   See Also: /tmp/af.449336a
19:33:10  Rollback error 126
19:33:10  Assert Failed: Rollback error 126
19:33:10  IBM Informix Dynamic Server Version 9.40.FC7
19:33:10   Who: Session(16, informix@zjnpmdb, 0, 5053c887
                Thread(97, fast_rec, 505386da0, 1)
                File: rsextlog.c Line: 1642
19:33:10   Results: Log record (OLDRSAM:HDELETE) in log 221077, offset 0x52701b4 was not rolled back
19:33:10   Action: Use 'onlog' to view the transaction and repair manually.
19:33:10  stack trace for pid 23350 written to /tmp/af.449336a
19:33:10   See Also: /tmp/af.449336a
19:33:21  Rollback error 126
19:33:21  Assert Failed: Dynamic Server must abort
19:33:21  IBM Informix Dynamic Server Version 9.40.FC7
19:33:21   Who: Session(16, informix@zjnpmdb, 0, 5053c887
                Thread(97, fast_rec, 505386da0, 1)
                File: rslog.c Line: 3511
19:33:21   Results: Dynamic Server must abort
19:33:21   Action: Reinitialize shared memory
19:33:21  stack trace for pid 23350 written to /tmp/af.449336a
19:33:21   See Also: /tmp/af.449336a
19:33:32  rslog.c, line 3511, thread 97, proc id 23350, Dynamic Server must abort.

lansworld

逻辑日志能不能跳过这段出错的内容? 或者直接清除掉所有的错误日志?

使用的平台如下：

主机： SUN 4800
存储： SUN 6120 磁盘阵列

informix 9.40.FC7

lansworld

这个问题昨天晚上已经解决，是找IBM工程师解决的，通过他们的一个工具把日志中错误的部分删除后恢复正常。