Hi everyone, When I read the source code of lustre, I have some confusions about those fids. I find that fid is the ino of inode most of time. Will it be changed When any operations of file objects be conducted? And why there are two fid(fid1, fid2) in the data structures like mds_body? thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.clusterfs.com/pipermail/lustre-devel/attachments/20070406/8937ffb6/attachment.html