A colleague said the day before that one of the servers that had an AIX RAC restarted the operating system, the resources of the cluster resource CSSD are always in the start state, and check the log output for the following: [ cssd][1286 ]clssnmvdiskverify:successful discovery of 0 disks[ cssd][1286]clssnmcompleteinitvfdiscovery:completing Initial voting file discovery[ cssd][1286]clssnmvfindinitialconfigs:no voting files found[  CSS D][1286] (: CSSNM00070:) clssnmcompleteinitvfdiscovery:voting file not found. Retrying discovery in seconds [ cssd][1029]clssscselect:cookie accept request 111272110[ &NBS P Cssd][1029]clssscevtypshrcon:getting client with Cmproc 111272110[ cssd][1029]clssgmregisterclient:proc (3/111272110), client (56/110a1a770) Log information from the cluster can be seen, due to the lack of scanning to the CRS of the relevant disk information, resulting in an abnormal result. cd/dev ls-l Discovery can see the relevant rhdiskpower disk Lsdev -CC disk discovers that the status of the associated drive letter is in the defined state, which identifies the disk before it is recognized, but is not currently available. This can be used to determine that the RAC cluster-related resources could not be read due to storage aspectsTo the contents of the OCR disk. Storage engineer Process: 1, check to verify that there are no problems with the physical line.  2, will delete the relevant identified disk on the small machine. rmdev-dl 3, remove all related links. Emcpower Delete dev=all 4, re-scan the drive letter. 5, adjust the sequence number of the re-scanned Hdiskpowern according to the LUN number of the normal machine.  6, save configuration. POWERMT save 7, re-check the status of the disk. LSDEV-CC Dev ensures that the status is available. After processing the storage problem, restart the machine and observe the cluster resource status.
