I'm trying to recover a Bareos server on new hardware AND new OS and am having
some difficulty.
(The exact problem I'm having is described below, but I'm also describing how &
why I got there; skip ahead if you want.)
Starting point:
* Bareos unknown version (nightly build from 15 era?) running on CentOS 6
* Yum repo misconfigured to point to -nightly builds
* PERC RAID with battery-backed write cache
* server was not on UPS, because I literally can't squeeze another UPS into
that room (and I thought there was no need to do backups during a power outage
anyway)
The triggering event:
* Momentary power outage
What has happened:
* RAID controller battery turns out to be unreliable, time to retire this
server immediately.
* Migrated server to newer hardware ("dd | nc" over the network for the entire
3TB disk, boot sectors and all)
* ran "yum upgrade" since I realized the kernel etc. were crazy out-of-date...
this pulled in Bareos 18-nightly without my realizing it
* eventually figured this out and rolled back to 17.2 (official RPMs)
* discovered there's no bareos-webui for CentOS 6 now
* backed up root filesystem
* reinstalled CentOS 7 from scratch
* reconfigured Bareos, merging previous config with 17.2 default config (it
appears to be configured correctly now)
...this is the point at which I realize I haven't seen a BSR file in months.
:-(
But, wait, my config backs up the Catalog to disk, I should be able to bscan
those on-disk volumes and get back to where I was before the power interruption.
Except I have far fewer files on disk than I should, and bscan insists on
restoring/updating the Media and Pool but not the Job or File or RestoreObject
when I scan them.
Now I'm starting to bscan the tapes that are still in the library, and they're
doing the same thing - updating the Media and Pool records, but no Jobs or
Files or RestoreObjects.
The most recent bscan command I've used was:
"sudo -u bareos bscan -B postgresql -c /etc/bareos -m -s -v -V 000016L6
/dev/tape/by-id/*nst"
The output from bscan looked like this:
***
bscan: bscan.c:779-0 Could not find Job for SessId=7 SessTime=1525554215 record.
bscan: bscan.c:779-0 Could not find Job for SessId=7 SessTime=1525554215 record.
bscan: bscan.c:779-0 Could not find Job for SessId=7 SessTime=1525554215 record.
bscan: bscan.c:1248-0 Created FileSet record ""
bscan: bscan.c:708-0 Could not find SessId=7 SessTime=1525554215 for EOS record.
08-May 10:09 bscan JobId 0: End of Volume at file 6 on device
"tapedrive-tl1000" (/dev/tape/by-id/scsi-350016977299e1010-nst), Volume
"000016L6"
bscan (0): bscan.c:382-0 ========== JobId=0 ========
bscan: bscan.c:1166-0 Updated Media record at end of Volume: 000016L6
08-May 10:09 bscan JobId 0: End of all volumes.
bscan: bscan.c:1166-0 Updated Media record at end of Volume: 000016L6
bscan: bscan.c:767-0 End of all Volumes. VolFiles=6 VolBlocks=0
VolBytes=268,721,467,884
Records added or updated in the catalog:
1 Media
1 Pool
0 Job
0 File
0 RestoreObject
***
There were a LOT of the "Could not find Job" errors; does that just mean I'm
scanning the wrong tape, or in the wrong order, or is there something more
fundamentally wrong?
I have:
* one BSR file on disk from April 28th that references an on-disk catalog backup
* a PgSQL backup from May 7th
* another (302-line!) BSR file on disk from May 7th that references some tapes
currently in the library
...but I don't know that I can trust any of these.
Any pointers/help/cluebats appreciated.
-Adam
--
You received this message because you are subscribed to the Google Groups
"bareos-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
For more options, visit https://groups.google.com/d/optout.
