condor_schedd hangs on a filelock in MRG 1.2
Issue
- Condor_schedd hangs on __fcntl for a log in NFS share
- Following stack is observed in logs:
Stack dump for process 18089 at timestamp 1281486087 (18 frames)
condor_schedd(dprintf_dump_stack+0x4e)[0x54846e]
condor_schedd[0x54a242]
/lib64/libpthread.so.0[0x3dd7a0eb10]
/lib64/libpthread.so.0(__fcntl+0x4f)[0x3dd7a0da6f]
condor_schedd(lock_file_plain+0x5b)[0x54b06b]
condor_schedd(lock_file+0x1f)[0x54b0ef]
condor_schedd(_ZN8FileLock6obtainE9LOCK_TYPE+0x6c)[0x554e7c]
condor_schedd(_ZN12WriteUserLog12doWriteEventEP9ULogEventbbP7ClassAd+0x6a)[0x56aaaa]
Environment
- MRG 1.2
- Log filesystem is on NFS
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.