Description of problem: I have a 4+2 ec volume, and a fault inject test (reboot a ec node every 600s), nfs client runs the fs_racer test of ltp. The ganesha.nfsd prints some logs in ganesha.log, ganesha.nfsd-19199[work-8] posix2fsal_type :FSAL :WARN :Unknown object type: 0 and logs in ganesha-gfapi.log, [dht-selfheal.c:668:dht_selfheal_dir_xattr_cbk] 0-test1-dht: layout setxattr failed on test1-disperse-0, path:/nfsshare/ltp-Fa5fNeMmWI/race/14/14 gfid:c51c4be6-b9a4-4cc0-96fd-f069018bf9d7 [No such file or directory] When meeting error, dht_selfheal_dir_xattr_cbk just return 0 to up layer that causes gfapi return success at mkdir with empty stat. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Will get back on this after reproducing this issue. Thanks for reporting. Susant
This bug is moved to https://github.com/gluster/glusterfs/issues/995, and will be tracked there from now on. Visit GitHub issues URL for further details