Below listed are rge buckets shown as directories on nfs mount point. stat always show size and blocks as 0 though it has data File: ‘missed-hellonew-another-newad-bucket-creation’ Size: 0 Blocks: 0 IO Block: 1048576 directory Device: 27h/39d Inode: 7556023705472190679 Links: 3 Access: (0777/drwxrwxrwx) Uid: ( 0/ root) Gid: ( 0/ root) Context: system_u:object_r:nfs_t:s0 Access: 1970-01-01 00:00:00.000000000 +0000 Modify: 1970-01-01 00:00:00.000000000 +0000 Change: 1970-01-01 00:00:00.000000000 +0000 Birth: - File: ‘nfsbucket’ Size: 0 Blocks: 0 IO Block: 1048576 directory Device: 27h/39d Inode: 8564301539577355726 Links: 3 Access: (0777/drwxrwxrwx) Uid: (4294967294/ UNKNOWN) Gid: (4294967294/ UNKNOWN) Context: system_u:object_r:nfs_t:s0 Access: 1970-01-22 02:03:28.000000000 +0000 Modify: 2016-07-23 13:11:23.000000000 +0000 Change: 2016-07-23 13:11:23.000000000 +0000 Birth: - File: ‘test’ Size: 0 Blocks: 0 IO Block: 1048576 directory Device: 27h/39d Inode: 16759487093327469849 Links: 3 Access: (0755/drwxr-xr-x) Uid: (4294967294/ UNKNOWN) Gid: (4294967294/ UNKNOWN) Context: system_u:object_r:nfs_t:s0 Access: 2016-07-23 19:55:45.000000000 +0000 Modify: 2016-07-23 19:55:45.000000000 +0000 Change: 2016-07-23 19:55:45.000000000 +0000 Birth: -
The RGW NFS implementation currently does not make any attempt to map bucket loading to directory size. The implementation does not attempt to calculate or estimate link counts for ordinary directories, either. Either of these might have updated behavior in 2.1, but may always have divergent behavior from ordinary NFS. RGW NFS is not expected to have fully consistent NFS semantics. We are still exploring what semantics best fit the needs of customers for NFS-S3 integration.
Updating the QA Contact to a Hemant. Hemant will be rerouting them to the appropriate QE Associate. Regards, Giri