On 2018-09-28 09:32 AM, Martin Simmons wrote:
On Thu, 27 Sep 2018 20:57:35 -0400, Gary Dale said:
On 2018-09-27 11:05 AM, Martin Simmons wrote:
On Wed, 26 Sep 2018 23:38:39 -0400, Gary Dale said:
On 2018-09-26 11:33 AM, Bill Arlofski wrote:
On 09/26/2018 06:22 AM, Martin Simmons wrote:
Try
setdebug level=150 trace=1 Storage
and look for a file matching *.trace (probably in the SD's WorkingDirectory).
...and you can also add options=t to have timestamps in the *.trace file.
Best regards,
Bill
I was waiting for the label command to fail again. It didn't take long.
The bconsole messages were the same. The trace doesn't tell me anything
new. Hopefully you understand it better than I. Here's the output from
the trace:
#cat /var/lib/bacula/bacula-sd.trace
bacula-sd: bnet.c:566-0 who=client host=192.168.1.17 port=9103
bacula-sd: dircmd.c:178-0 Got a DIR connection at 26-Sep-2018 23:30:25
bacula-sd: cram-md5.c:69-0 send: auth cram-md5 challenge
<56845663.1538019025@bacula-sd> ssl=0
bacula-sd: cram-md5.c:133-0 cram-get received: auth cram-md5
<554803286.1538019025@bacula-dir> ssl=0
bacula-sd: cram-md5.c:157-0 sending resp to challenge: cH+vQ7+Fdw/6P7+9R7/jRA
bacula-sd: dircmd.c:204-0 Message channel init completed.
bacula-sd: dircmd.c:806-0 Found device FileStorage
bacula-sd: dircmd.c:850-0 Found device FileStorage
bacula-sd: block_util.c:143-0 New block len=64512 block=7f19c8001160
bacula-sd: acquire.c:673-0 Attach 0xc8001c78 to dev "FileStorage"
(/home/shares/backup)
bacula-sd: vol_mgr.c:657-0 find_read_vol: read_vol_list empty.
bacula-sd: vol_mgr.c:851-0 Vol=weekly-backup-pool-37 not in use.
bacula-sd: vol_mgr.c:381-0 enter reserve_volume=weekly-backup-pool-37
drive="FileStorage" (/home/shares/backup)
bacula-sd: vol_mgr.c:286-0 new Vol=weekly-backup-pool-37 slot=0 at 7f19c8011e88
dev="FileStorage" (/home/shares/backup)
bacula-sd: vol_mgr.c:547-0 set in_use. vol=weekly-backup-pool-37
dev="FileStorage" (/home/shares/backup)
bacula-sd: vol_mgr.c:581-0 Inc walk_start use_count=2
volname=weekly-backup-pool-37
bacula-sd: vol_mgr.c:220-0 List end new volume: weekly-backup-pool-37 in_use=1 swap=0
slot=0 on file device "FileStorage" (/home/shares/backup)
bacula-sd: vol_mgr.c:693-0 Clear in_use vol=weekly-backup-pool-37 slot=0
bacula-sd: vol_mgr.c:710-0 set not reserved vol=weekly-backup-pool-37 slot=0 writers=1
reserves=0 dev="FileStorage" (/home/shares/backup)
bacula-sd: vol_mgr.c:740-0 Clear in_use vol=weekly-backup-pool-37 slot=0
bacula-sd: vol_mgr.c:746-0 Remove volume weekly-backup-pool-37 slot=0
dev="FileStorage" (/home/shares/backup)
bacula-sd: mount.c:384-3298 Again dir_find_next_append...
bacula-sd: mount.c:374-3298 No appendable volume. Calling wait_for_device
bacula-sd: wait.c:270-3298 Enter wait_for_device. busy=1 dcrvol= devvol=
bacula-sd: mount.c:384-3298 Again dir_find_next_append...
bacula-sd: mount.c:374-3298 No appendable volume. Calling wait_for_device
bacula-sd: wait.c:270-3298 Enter wait_for_device. busy=1 dcrvol= devvol=
bacula-sd: mount.c:384-3298 Again dir_find_next_append...
bacula-sd: mount.c:374-3298 No appendable volume. Calling wait_for_device
bacula-sd: wait.c:270-3298 Enter wait_for_device. busy=1 dcrvol= devvol=
Is it still printing
3934 Device ""FileStorage" (/home/shares/backup)" is being initialized.
when the label command fails? If so, then the output of the "status storage"
command might give more information.
__Martin
That message always prints just before the label command fails. It makes
no sense to me because what is there to initialize? It's a folder on a
disk (/home/shares/backup).
That message means the FileStorage device is trying to "initialize", i.e. find
a volume.
I just got an example of another one:
Connecting to Storage daemon File at
molar.dentaloffice.rossland.dental:9103 ...
Sending label command for Volume "weekly-backup-pool-40" Slot 0 ...
3908 Error reserving volume: vol_mgr.c:419 Cannot free Volume
"weekly-backup-pool-37", because it is reserved by someone else.
Label command failed for Volume weekly-backup-pool-40.
You can't label a new volume while a running job is actively writing to
another volume.
Maybe that is also the cause of the original problem? In general, you have to
either label volumes when nothing is running or label one when Bacula asks you
to.
__Martin
That makes sense. Unfortunately Bacula doesn't seem to want to ask me.
It just sits there waiting for a volume that it won't let me label...
I've left myself logged in and running bconsole to see if that helps.
Hopefully it will issue the message to the console and wait for me to
label a new volume.
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
