|
Hi, I’m currently trying to restore a filesystem on a production
Linux PostgreSQL database server. I’ve been doing weekly full
backups with daily incrementals for some time now using an Exabyte 480
autoloader. The files to be restored span 4 tapes. The restore
process was going smoothly until it hit the end of the first tape. As
expected, it reached the end of the tape and detected the EOF: portal-fd: -rw------- 1 postgres
postgres1073741824 2006-01-30 07:01:28
/usr2/pgsql-8.1.1/data/base/16390/266709 Then it restored a few more files (or was it just the log
entries were out of order – doesn’t make sense to me...?): portal-fd: -rw------- 1 postgres postgres 8192
2006-01-23 11:08:00 /usr2/pgsql-8.1.1/data/base/16390/266719 ... ... Then it detected the end-of-volume on the tape: 08-Feb 14:05 linuxapp1-sd: End of Volume at file 27 on
device "Drive-1" (/dev/nst0), Volume "000020" Then restored a few more files: portal-fd: -rw------- 1 postgres
postgres 8192 2005-12-19 11:35:37
/usr2/pgsql-8.1.1/data/base/16390/266899 <snip> ... Then tried to switch to the second tape to restore from: 08-Feb 14:05 linuxapp1-sd: 3301 Issuing autochanger
"loaded drive 0" command. That’s when it got this error: 08-Feb 14:08 linuxapp1-sd:
RestoreFiles.2006-02-08_13.22.45 Warning: acquire.c:230 Could not unserialize
Volume label: ERR=label.c:720 Expecting Volume Label, got FI=333012
Stream=SPARSE-DATA len=12476 After searching the email list archives, I’m currently
trying to do the restore using the “-p” option on bacula-sd, but so
far it’s not working either. I’m running bacula 1.38.2 on a
Linux system. Please help! -Chris |
- [Bacula-users] Urgent: Error during restore -- "Could n... Chris Wilson
