Another glitch

Ben Escoto
Sun, 24 Feb 2002 12:35:36 -0800

Content-Type: text/plain; charset=us-ascii

>>>>> "docv" == docv  <>
>>>>> wrote the following on Sat, 23 Feb 2002 11:05:04 -0600

  docv> Ok, I started rdiff-backup around 1:40 am this morning and
  docv> when I checked on it around 10:00 am the HD on the source box
  docv> was going crazy (making all kinds of activity noise) and the
  docv> gnome terminal I ran rdiff-backup from was still waiting for
  docv> it to complete. There's less than 1.5 gig of data on the
  docv> source and I know it shouldn't take that long to backup.

  Ben> I've never seen this...

Actually, someone's email recently reminded me that I have seen this
before, but it isn't an rdiff-backup bug strictly speaking..  If you
try to backup a directory into itself then rdiff-backup will loop
forever.  For instance, if you type

rdiff-backup foo foo/backup

then rdiff-backup will create foo/backup, and start backing files up
into it.  But then it will find foo/backup (which is in the foo
directory after all), and try to back that up into foo/backup/backup.
And then foo/backup/backup/backup, etc.

    In general these cycles can't all be detected (you might be
running rdiff-backup remotely with system NFS mounted on the other)
but I suppose it wouldn't hurt to try to detect warn about these cases
when they occur locally.

    Anyway, this may not be your problem, but now that I think about
it, it probably is..

Ben Escoto

Content-Type: application/pgp-signature

Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Exmh version 2.5 01/15/2001