The problem could only occur when a GPFS client has a file system panic at the same time as the new file system manager is performing a take over after the old manager resigned. Fix a signal 11 problem in multi-cluster environment when gpfs daemon relay the fsync request through metanode ...