[NCLUG] discrepency between reiserfs and ext3?

[jbass at dmsd.com]

"Robie Lutsey" <robiel at tgstech.com> writes:
> [Robie ] Do different file-systems reply to calls like ls differently?
> For example if I did an ls on ext3:/home and reiserfs:/home, I get the
> same thing.  But are the intermediate steps the same?  Are there other
> calls that would return the same information in a different order, all
> else being equal?

Shouldn't, at least for those POSIX attributes that are visible over NFS.

Well, having ruled out data corruption and basic attributes, that pretty
much covers 99% of what a POSIX compliant application can see over NFS.
There are a few other visible fields for a file: mtime, ctime, ... etc
which are in the stat struct returned by a stat/fstat system call.

              struct stat {
                  dev_t         st_dev;      /* device */
                  ino_t         st_ino;      /* inode */
                  mode_t        st_mode;     /* protection */
                  nlink_t       st_nlink;    /* number of hard links */
                  uid_t         st_uid;      /* user ID of owner */
                  gid_t         st_gid;      /* group ID of owner */
                  dev_t         st_rdev;     /* device type (if inode device) */
                  off_t         st_size;     /* total size, in bytes */
                  blksize_t     st_blksize;  /* blocksize for filesystem I/O */
                  blkcnt_t      st_blocks;   /* number of blocks allocated */
                  time_t        st_atime;    /* time of last access */
                  time_t        st_mtime;    /* time of last modification */
                  time_t        st_ctime;    /* time of last change */
              };

A truly paranoid application could trigger checks off any or all of
these fields.

Past that, there aren't any other attributes I can think of that are
visible on the other side of an NFS connection, that would be changed
by simply copying the file between the two filesystems.

John Bass