Hi Jean

As I mentioned in a previous mail this week, my problems miraculously
disappeared when I upgrade the client to 2.5.2p1, still touching wood
about that...

I'd believed the problem was relate to 

>From the sendsize debug log:

sendsize[1907]: time 24.758: child 1911 terminated with signal 4

Which I'd seen mentioned somewhere as a linux kernel process signal
handling change, hacked all binary outcome checks to include and convert
the signal '4' to something sensible in sendmail and sendbackup source
code. recompiled but to no avail. Then the upgrade cornered the goose :)

I still think that signal 4 is/was guilty.

Cheers

Marc O'Brien

----- Original Message -----
From: Jean-Francois Malouin <Jean-Francois.Malouin AT bic.mni.mcgill DOT ca>
Date: Thursday, January 17, 2008 7:41 pm
Subject: more of the dreaded 'missing size line from sendbackup'
To: AMANDA users <amanda-users AT amanda DOT org>

> Hello again,
> 
> Yet another case of the driver reporting 'missing size line from
> sendbackup' just before the DLE (~1TB) was entirely in the holddisk...
> 
> Here's what the server (2.5.2p1) logs show: 
> 
> log:
> 
> FAIL dumper yorick /data/nih/nih1 20080116 0 [missing size line 
> from sendbackup]
>  sendbackup: start [yorick:/data/nih/nih1 level 0]
>  sendbackup: info BACKUP=/usr/freeware/bin/tar
>  sendbackup: info RECOVER_CMD=/usr/freeware/bin/tar -xpGf - ...
>  sendbackup: info end
>  ? dumper: strange [missing size line from sendbackup]
> PARTIAL chunker yorick /data/nih/nih1 20080116 0 [sec 49050.184 kb 
> 933151530 kps 19024.4]
> 
> amdump:
> 
> driver: state time 50745.505 free kps: 1003626 space: 979020896 
> taper: idle idle-dumpers: 11 qlen tapeq: 0 run q: 0 roomq: 0 
> wakeup: 0 driver-idle: no-dumpers
> driver: interface-state time 50745.505 if default: free 1002226 if 
> local: free 1000 if le0: free 400
> driver: hdisk-state time 50745.505 hdisk 0: free 979020896 dumpers 1
> driver: result time 50745.522 from dumper1: FAILED 01-00002 
> "[missing size line from sendbackup]"
> driver: send-cmd time 50745.522 to chunker1: FAILED 01-00002
> driver: state time 50745.522 free kps: 1003626 space: 979020896 
> taper: idle idle-dumpers: 11 qlen tapeq: 0 runq: 0 roomq: 0 wakeup: 
> 0 driver-idle: no-dumpers
> driver: interface-state time 50745.522 if default: free 1002226 if 
> local: free 1000 if le0: free 400
> driver: hdisk-state time 50745.522 hdisk 0: free 979020896 dumpers 1
> driver: result time 50745.522 from chunker1: PARTIAL 01-00002 
> 933151530 "[sec 49050.184 kb 933151530 kps 19024.4]"
> driver: finished-cmd time 50745.522 chunker1 chunked 
> yorick:/data/nih/nih1
> On the client (2.5.2p1), the amandad/sendbackup/runtar debug files:
> 
> amandad.20080116223817.debug:
> 
> amandad: time 49049.651: security_stream_seterr(1003ca70, write 
> error to :
> Resource temporarily unavailable)
> amandad: time 49049.663: sending NAK pkt:
> <<<<<
> ERROR write error on stream 499999: write error to : Resource 
> temporarilyunavailable
> >>>>>
> 
> sendbackup.20080116223817.debug:
> 
> sendbackup-gnutar: time 724.655: /opt/amanda/av24-2/libexec/runtar: 
> pid 7955181
> sendbackup: time 724.656: started backup
> sendbackup: time 49049.606: index tee cannot write [Broken pipe]
> sendbackup: time 49049.616: pid 7963340 finish time Thu Jan 17 
> 12:15:46 2008
> 
> 
> runtar.20080116225021.debug:
> 
> runtar: debug 1 pid 7955181 ruid 666 euid 0: start at Wed Jan 16 
> 22:50:21 2008
> runtar: time 0.001: version 2.5.2p1
> /usr/freeware/bin/tar version: tar (GNU tar) 1.13.25
> 
> config: left2
> runtar: debug 1 pid 7955181 ruid 0 euid 0: rename at Wed Jan 16 
> 22:50:21 2008
> running: /usr/freeware/bin/tar: 'gtar' '--create' '--file' '-' '--
> directory''/data/nih/nih1' '--one-file-system' '--listed-incremental'
> '/opt/amanda-av24-2/var/amanda/left2/gnutar-
> lists/yorick_data_nih_nih1_0.new''--sparse' '--ignore-failed-read' 
> '--totals' '.' 
> runtar: time 0.017: pid 7955181 finish time Wed Jan 16 22:50:21 2008
> 
> 
> Looks like another goose chase to me...
> Any ideas on how to debug and track this issue?
> 
> regards,
> jf
> -- 
> <° ><
>