RAC exceptions caused by storage link problems. The RAC exception caused by storage link problems was added to me yesterday, saying that the RAC problem was encountered. The environment is oracle11.0.2.4 and 2 nodes. As shown in the following figure, the cluster will not automatically switch when the CRSD is down. RAC exceptions caused by storage link problems
Yesterday, someone added me and said that I encountered a RAC problem. The environment is oracle 11.0.2.4 and 2 nodes. As shown in the following figure, the cluster will not automatically switch when the CRSD is down. It appears several times.
I asked this guy to provide logs.
Through the log, we can see that the REBOOT of the CRSD service on nodes 1 and 2 is only one second, and the Voting file/dev/raw/raw1 cannot provide services for only two seconds. CRSD restarts at the same time at two nodes, and it can be determined that it is not a RAC issue.
Node 23:25:38-03-18. 643: [cssd (6930)] CRS-1615: No I/O has completed after 50% of the maximum interval. voting file/dev/raw/raw1 will be considered not functional in 99220 milliseconds
23:26:58 crsd rebootcrsd exiting: cocould not init OCR, code: 26
Node 23:25:36. 834: [cssd (6590)] CRS-1615: No I/O has completed after 50% of the maximum interval. voting file/dev/raw/raw1 will be considered not functional in 99760 milliseconds
23:26:59 crsd rebootcrsd exiting: cocould not init OCR, code: 26
Looking at the OS log, the connectivity of the remote optical fiber port (WWN = 50: 0b: 34: 20: 0f: f4: f8: 08) is lost. It must have been a problem with the storage link. Node 1Mar 18 23:26:38 xxxxr01 kernel: bfa 0000: 81: 00.0: Remote port (WWN = 50: 0b: 34: 20: 0f: f4: f8: 08) connectivity lost for logical port (WWN = 21: 00: 00: 24: ff: 09: a5: 8a) Mar 18 23: 26: 38xxxxr01kernel: bfa 0000: 81: 00.0: target (WWN = 50: 0b: 34: 20: 0f: f4: f8: 08) connectivity lost for initiator (WWN = 21: 00: 00: 24: ff: 09: a5: 8a)
Node 2Mar 18 23:26:48 xxxx02 kernel: bfa 0000: 81: 00.0: Remote port (WWN = 50: 0b: 34: 20: 0f: f4: f8: 08) connectivity lost for logical port (WWN = 21: 00: 00: 24: ff: 09: a5: a2) Mar 18 23:26:48 xxxr02 kernel: bfa 0000: 81: 00.0: target (WWN = 50: 0b: 34: 20: 0f: f4: f8: 08) connectivity lost for initiator (WWN = 21: 00: 00: 24: ff: 09: a5: a2)
Today, I chatted with this buddy and I have contacted the storage vendor to check the storage link. Storage is from Hongshu. I first heard of Chinese-made storage, which is quite informative.
In RAC, there are problems, and most of the time it is not RAC's own problems. I have encountered a problem with the storage link that causes RAC to be unable to provide services several times. When a DBA encounters a RAC exception, both the software and hardware should be checked to quickly locate the problem.
Reprinted please note:
Http://blog.chinaunix.net/uid/23284114.html
Yesterday, someone added me and said that I encountered a RAC problem. The environment is oracle 11.0.2.4 and 2 nodes. As shown in the following figure, the cluster will not automatically switch when the CRSD is down. And...
