* Jean-Louis Martineau <jmartin...@carbonite.com> [20170220 14:24]: > On 20/02/17 11:47 AM, Jean-Francois Malouin wrote: > > Hi, > > > > Attempting to restore a DLE with amfetchdump with: > > > > su amanda -c "~amanda/sbin/amfetchdump -p -d tape:/dev/nst0 top localhost > > /data_/mica1 20170219 " | tar -xpGf - > > > > returns 'Error writing to fd 7: Broken pipe'. The logfile shows: > Assuming fd 7 is the pipe to tar, it means that tar exited before > amfetchdump send all the data, do you get any error from tar? > You can try to do it in 2 step: > > $ ~amanda/sbin/amfetchdump -p -d tape:/dev/nst0 top localhost /data_/mica1 > 20170219 > datafile > > Do the data file contain the backup? ls -l, file datafile > > $ tar -xpGf datafile > > Check system/selinux logs for error.
Nothing relevant shows up in the host's syslog. Doing the restore in 2 steps as proposed by Jean-Louis terminates without an error: ~# pwd /holddisk/restore ~# su amanda -c "~amanda/sbin/amfetchdump -p -d /dev/nst0 top localhost /data_/mica1 20170219" > mica1.datafile dow:/holddisk/restore# echo $? 0 ~# tar -C /holddisk/restore/mica1 -xpGf mica1.datafile ~# echo $? 0 The amfetchdump debug file shows all is good. I know I had 2 issues in the same post: what about the 2nd one, amrestore not doing recursive restore? thanks, jf > > Jean-Louis > > > > > Mon Feb 20 11:26:03.128402078 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > xfer_cancel_with_error: Error writing to fd 7: Broken pipe > > Mon Feb 20 11:26:03.128432851 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > xfer_queue_message: MSG: <XMsg@0x7fcf902015a0 type=XMSG_ERROR > > elt=<XferElementGlue@0x317d000> version=0> > > Mon Feb 20 11:26:03.128457912 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > xfer_queue_message: MSG: <XMsg@0x7fcf90201610 type=XMSG_CANCEL > > elt=<XferSourceRecovery@0x2e46000> version=0> > > Mon Feb 20 11:26:03.128835292 2017: pid 14583: thd-0x1502000: amfetchdump: > > Cancelling <Xfer@0x3179980 (<XferSourceRecovery@0x2e46000> -> > > <XferDestFd@0x317c810>)> > > Mon Feb 20 11:26:03.128867340 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > sending XMSG_CRC message 0x317c810 > > Mon Feb 20 11:26:03.128878700 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > pull_and_write CRC: f9201a62 size 102900957184 > > Mon Feb 20 11:26:03.128886874 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > xfer_queue_message: MSG: <XMsg@0x7fcf90201680 type=XMSG_CRC > > elt=<XferDestFd@0x317c810> version=0> > > Mon Feb 20 11:26:03.128900594 2017: pid 14583: thd-0x2e49c00: amfetchdump: > > xfer_queue_message: MSG: <XMsg@0x7fcf90201760 type=XMSG_DONE > > elt=<XferElementGlue@0x317d000> version=0> > > Mon Feb 20 11:26:03.129525236 2017: pid 14583: thd-0x1502000: amfetchdump: > > dest_crc: f9201a62:102900957184 > > Mon Feb 20 11:26:03.129946543 2017: pid 14583: thd-0x1502000: amfetchdump: > > /opt/amanda/perl/Amanda/Restore.pm:1892:info:4900012 100491264 kb > > Mon Feb 20 11:26:03.130359016 2017: pid 14583: thd-0x1502000: amfetchdump: > > /opt/amanda/perl/Amanda/Restore.pm:1920:error:4900055 Error writing to fd > > 7: Broken pipe > > Mon Feb 20 11:26:03.130501339 2017: pid 14583: thd-0x1502000: amfetchdump: > > /opt/amanda/perl/Amanda/Restore.pm:2137:error:4900068 Error writing to fd > > 7: Broken pipe > > > > What is the error all about? > > > > A previous amrecover run seems to have gone to all the motions > > correctly, but in the end the restore done in temp dir yield > > nothing...an empty directory... > > > > Any hint? > > Thanks, > > jf
