Unable to rdump -L -0 a 20G fs?

[Alexander Langer]

Hi!

I have a problem with dump (rdump), as it won't dump -0 my data
filesystem.  Higher dumplevels (such as -1 - -9) do work.

It just seems to hang forever:
root at cure ~ # RSH=ssh /sbin/rdump -auL -b 1000 -C 32 -0 -f
	defiant-local:/backups/cure/cure-data-level0.dump /data
  DUMP: Connection to defiant-local established.
IP_TOS:IPTOS_THROUGHPUT setsockopt: Operation not supported
TCP_NODELAY setsockopt: Operation not supported
  DUMP: Date of this level 0 dump: Tue Feb  3 18:31:02 2004
  DUMP: Date of last level 0 dump: the epoch
  DUMP: Dumping snapshot of /dev/da0s1f (/data) to
	/backups/cure/cure-data-level0.dump on host defiant-local
  DUMP: mapping (Pass I) [regular files]
  DUMP: Cache 32 MB, blocksize = 65536
  DUMP: mapping (Pass II) [directories]
  DUMP: estimated -1529814507 tape blocks.
  DUMP: dumping (Pass III) [directories]

This already hangs for 12 hours or so, after I killed the last -0 dump
which already lasted 3 days (!) being in pass 3 (which is rather bad if
you want to do daily backups) The last level 1 dump took only 6.5 hours.

Here's the output of top(1):
  PID USERNAME PRI NICE   SIZE    RES STATE  C   TIME   WCPU    CPU COMMAND
88902 root      98    0 39220K 34396K RUN    2  17.0H 92.58% 92.58% rdump

root at cure ~ # ps ax | grep dump
88875  p4  I+     0:00.40 /sbin/rdump -auL -b 1000 -C 32 -0 -f
		defiant-local /data (rdump)
88901  p4  I+     0:00.03 rdump: /dev/da0s1f: pass 3: directories (rdump)
88902  p4  R+   1022:04.11 /sbin/rdump -auL -b 1000 -C 32 -0 -f
		defiant-local /data (rdump)
88903  p4  I+     0:00.00 /sbin/rdump -auL -b 1000 -C 32 -0 -f
		defiant-local /data (rdump)
88904  p4  I+     0:00.00 /sbin/rdump -auL -b 1000 -C 32 -0 -f
		defiant-local /data (rdump)

root at cure ~ # df -ih /data
Filesystem    Size   Used  Avail Capacity iused   ifree %iused  Mounted on
/dev/da0s1f    20G   5.9G    12G    33%   93544 2567830    4%   /data

/dev/da0s1f on /data (ufs, local, soft-updates) (UFS2, that is)

root at cure ~ # tunefs -p /data
tunefs: ACLs: (-a)                                         disabled
tunefs: MAC multilabel: (-l)                               disabled
tunefs: soft updates: (-n)                                 enabled
tunefs: maximum blocks per file in a cylinder group: (-e)  2048
tunefs: average file size: (-f)                            16384
tunefs: average number of files in a directory: (-s)       64
tunefs: minimum percentage of free space: (-m)             8%
tunefs: optimization preference: (-o)                      time
tunefs: volume label: (-L)                                 

FreeBSD cure.... 5.2-RELEASE FreeBSD 5.2-RELEASE #12: Sat Jan 17 11:39:48
CET 2004 alex at cure...:/usr/obj/usr/src/sys/CURE  i386

On the remote machine (a 4.9-RELEASE-p1, the dumpfile has filesize 0.
Usually, rdump writing data to the dumpfile from the very beginning,
but:
-rw-r--r--  1 root  wheel            0 Jan 15 12:24 cure-data-level0.dump
(Jan 15 is the day I "touched" the dumpfile).

Here is the level 1 dump.
-rw-r--r--  1 root  wheel  23431680000 Jan 22 06:27 cure-data-level1.dump

This seems to be a full-backup as well, as I've never succeeded in a
level 0 dump before.

Note that I get negative estimated tape blocks (and estimated time) for
level 1 dumps as well, but at least the dumps finish.

Anyone has an explanation?  Yes, there is enough diskspace available on
/backups on the remove-machine :-)

Ciao

Alex