On Oct 27, 2006, at 5:07 PM, Stephen Joyce wrote:

It looks like it actually performed full backups of some hosts (kw33,
kw109, kw124, kw9) normally between 1am and 3am, then linked them at 4am.
The BackupPC_nightly task took the next ~8 hours, during which no backups
occurred. Then, it tried to link the same hosts again...

I don't see anything that looks like an error in their logs aside from the usual vanishing files (that is normal).  The system is currently trying to back up some Windows servers (which always fail) and has all the 200+ servers scheduled.  We'll see what it has accomplished by tomorrow.  It did this the day before... had all 200+ hosts scheduled and then for some reason marked them all idle with nothing to do for no reason I can discern.

I'm assuming that the first BackupPC_link for those hosts failed for some
reason since it's trying again... can you show us the last bad xferlog for
one of those hosts, if it exists, or any lines in the plain xferlog that
don't look normal?

Here's one:

Contents of file /var/lib/backuppc/pc/kw33.internal.chinoc.net/XferLOG.6.z, modified 2006-10-27 01:51:48 (Extracting only Errors)

Running: /usr/bin/ssh -p 8889 -q -x -l backup kw33.internal.chinoc.net nice -n 19 sudo /var/backups/rsyncSend --server --sender --numeric-ids --perms --owner --group --devices --links --times --block-size=2048 --recursive --exclude=/proc --exclude=/sys --exclude=/home --exclude=/vol01 --exclude=/vol03 --exclude=/vol04 --exclude=/home/pluginfeeds --exclude=/var/lib/mysql --ignore-times . /
Xfer PIDs are now 16687
Got remote protocol 29
Negotiated protocol version 26
Sent exclude: /proc
Sent exclude: /sys
Sent exclude: /home
Sent exclude: /vol01
Sent exclude: /vol03
Sent exclude: /vol04
Sent exclude: /home/pluginfeeds
Sent exclude: /var/lib/mysql
Xfer PIDs are now 16687,16688
[ skipped 40972 lines ]
Done: 38648 files, 707511199 bytes

Oh, and backuppc is almost certainly not having trouble with the size of
your filesystem. I'm using it with a 9TB partition without problems. Are
you confusing your pool, which if you're using compression should be empty,
with your Cpool, which has ~1TB?

Yes, I realized this right after I sent the e-mail.  It was one of those d'oh moments.  :-)

Oh, also can you verify that you're not running out of inodes on your
filesystem? What's the output of "df -hi"?

jeward@kw157:~$ df -hi
Filesystem            Inodes   IUsed   IFree IUse% Mounted on
/dev/md1                8.8M     49K    8.7M    1% /
tmpfs                   127K       1    127K    1% /dev/shm
/dev/md0                 63K      31     63K    1% /boot
/dev/sdc                466M     15M    451M    4% /var/lib/backuppc