To use LABEL= for xfs mountpoints? Doesn't seem to be working for me. I formatted all my fs, then went back and did xfs_admin -L name /dev/sdN I then added LABEL=name /mountpoint xfs defaults 1 2 and
as a side note to this, I didn't make any partitions, I'm just using the whole disk. I dunno if that makes a difference but thought I'd ask. -- Austin Gonyou <austin@xxxxxxxxxxxxxxx> Coremetrics, Inc
First, how long a string, xfs has room for 14 chars, second, how recent is your mount command, it has to be able to understand xfs on disk format. Steve
Not sure, the mount command scans everything which appears in /proc/partitions if your filesystem is in there it should work, provided the items I just mentioned are taken care of. Steve
Gotcha. I'll check it out. :) -- Austin Gonyou <austin@xxxxxxxxxxxxxxx> Coremetrics, Inc. Attachment: signature.asc Description: This is a digitally signed message part
They are, the problem is mount explicitly skips them (see util-linux/mount/mount_by_label.c::uuid_cache_init...) ... for(s = ptname; *s; s++); if (isdigit(s[-1])) { ...[do good stuff]... } and right
There is no very good reason. Under Linux one customarily partitions disks, but there is no need to do so. Under BSD I see whole-disk filesystems more often. As soon as you file a bug report explain
The answer was already given, complete devices don't show up in /proc/partitions so the disk will not get scanned for mount by label. Just make one large partition. -Eric -- Seth It might just be you
Good deal then. :) -- Austin Gonyou <austin@xxxxxxxxxxxxxxx> Coremetrics, Inc. Attachment: signature.asc Description: This is a digitally signed message part