|
Hello,
On the software side, it is clear that the OS is sending back an
EOT status. About the only other thing to check before looking at
your tape wiring/controller ... is to be sure you have not put the
tape drive into some unusual mode such as BSD mode. This would be
something that is done via the OS and the default configuration
files. If you are just running a stock Linux system, what I just
mentioned about unusual modes is very unlikely to occur. The only
other software possibility that I can think of is who built your
SD binaries, with what options, and with which compiler.
At this point it still looks like something in the hardware but
that is just a guess.
Best regards,
Kern
On 05.02.2018 17:41, Tom Plancon wrote:
Hi Kern,
Yes, this has happened on several tapes at this point. They are
all brand new HP LTO-5 tapes.
No, I'm not using the Windows SD. The SD is running on the
Centos 7 machine which runs the tape drive.
So this AM I'm running the btape Fill routine. It it still
running and it's over 130Gb at this time, that's 6 times the
21Gb I was getting on the backup. Is it safe to say that the
controller is not the issue?
Thanks again for your help!
On 2/3/2018 3:58 AM, Kern Sibbald
wrote:
Hello,
In this case, Bacula received an EOT marker, so it stopped
writing. My best guess is that the tape is defective or has a
misplaced EOT marker. Does this happen on all your tapes?
If so, then there might be something wrong with your tape
controller.
You are not by any chance using the Windows SD are you?
(pretty unlikely, but just checking).
Another thing to do is to point btape at the volume (assuming
you can afford to overwrite the data) and give it the "fill"
command and see how much data btape can write.
Then try btape on a different tape to see what it does.
Best regards,
Kern
On 02.02.2018 22:53, Tom Plancon
wrote:
Kern,
Thanks much for the response! Here is the job output from
Bacula, there was nothing in /var/log/messages for the time
period.
01-Feb 17:01 godwit-bacula-dir JobId 33: No prior Full
backup Job record found.
01-Feb 17:01 godwit-bacula-dir JobId 33: No prior or
suitable Full backup found in catalog. Doing FULL backup.
01-Feb 17:01 godwit-bacula-dir JobId 33: Start Backup JobId
33, Job=HomeTape.2018-02-01_17.01.01_03
01-Feb 17:01 godwit-bacula-dir JobId 33: Using Device
"LTO-5" to write.
01-Feb 17:01 bacula-sd JobId 33: Wrote label to prelabeled
Volume "001" on tape device "LTO-5" (/dev/st0)
01-Feb 17:01 bka-design-fd JobId 33: Generate VSS snapshots.
Driver="Win64 VSS"
01-Feb 17:01 bka-design-fd JobId 33: Snapshot mount
point: E:\
01-Feb 17:45 bacula-sd JobId 33: Re-read of last block
succeeded.
01-Feb 17:45 bacula-sd JobId 33: End of medium on Volume
"001" Bytes=21,474,819,072 Blocks=332,880 at 01-Feb-2018
17:45.
01-Feb 17:47 godwit-bacula-dir JobId 33: Created new
Volume="Hme-0002", Pool="HomePool", MediaType="LTO-5" in
catalog.
01-Feb 17:47 bacula-sd JobId 33: Please mount append Volume
"Hme-0002" or label a new one for:
Job: HomeTape.2018-02-01_17.01.01_03
Storage: "LTO-5" (/dev/st0)
Pool: HomePool
Media type: LTO-5
<....tape request repeated....>
02-Feb 09:48 bacula-sd JobId 33: Fatal error: Job 33
canceled.
02-Feb 09:48 bacula-sd JobId 33: Elapsed time=16:47:45,
Transfer rate=354.8 K Bytes/second
02-Feb 09:48 bacula-sd JobId 33: Fatal error: askdir.c:355
NULL Volume name. This shouldn't happen!!!
02-Feb 09:48 godwit-bacula-dir JobId 33: Bacula
godwit-bacula-dir 7.0.5 (28Jul14):
Build OS: x86_64-redhat-linux-gnu redhat
Enterprise release
JobId: 33
Job: HomeTape.2018-02-01_17.01.01_03
Backup Level: Full (upgraded from Incremental)
Client: "bka-design-fd" 7.0.5 (04Aug14)
Microsoft Windows Server 2008 R2 Standard Edition Service
Pack 1 (build 76
01), 64-bit,Cross-compile,Win64
FileSet: "Home Set" 2018-02-01 17:01:01
Pool: "HomePool" (From Job resource)
Catalog: "MyCatalog" (From Client resource)
Storage: "LTO-5" (From Job resource)
Scheduled time: 01-Feb-2018 17:00:55
Start time: 01-Feb-2018 17:01:03
End time: 02-Feb-2018 09:48:54
Elapsed time: 16 hours 47 mins 51 secs
Priority: 10
FD Files Written: 19,340
SD Files Written: 19,373
FD Bytes Written: 21,301,085,904 (21.30 GB)
SD Bytes Written: 21,457,844,820 (21.45 GB)
Rate: 352.3 KB/s
Software Compression: None
VSS: no
Encryption: no
Accurate: no
Volume name(s): 001
Volume Session Id: 1
Volume Session Time: 1517515212
Last Volume Bytes: 0 (0 B)
Non-fatal FD errors: 0
SD Errors: 1
FD termination status: Canceled
SD termination status: Canceled
Termination: Backup Canceled
:
Thanks again.
On 2/2/2018 4:16 PM, Kern Sibbald
wrote:
Hello,
Unfortunately without the *complete* job output from Bacula
when "Bacula called the tape full", we cannot do much for
you.
You should also look and see if there are any messages in
the OS log at the time the tape was declared full. If there
are any please include them too.
Best regards,
Kern
On 02.02.2018 20:42, Tom Plancon wrote:
Hi folks,
Setting up Bacula 7.0.5 on Centos 7, backing up to a HPE
LTO-5 Ultrium 3000 external SAS drive. Everything going
fairly smoothly; bacula-sd finds the drive and btape-test
reports full success. However, tried doing a test backup
from a Win client using a directory of 134Gb. Backup
started and appeared to run normally, then after about
21Gb written to tape, Bacula called tape full and
requested a new tape be mounted!
I've seen discussions re: block sizes on LTO-5 devices,
but much of it a bit over my head. This is the first LT0-5
device I've used although had great success with LT0-4
devices. Here it the output of Tapeinfo:
Product Type: Tape Drive
Vendor ID: 'HP '
Product ID: 'Ultrium 5-SCSI '
Revision: 'Z6ED'
Attached Changer API: No
SerialNumber: 'HUJ6326M2G'
MinBlock: 1
MaxBlock: 16777215
SCSI ID: 0
SCSI LUN: 0
Ready: yes
BufferedMode: yes
Medium Type: Not Loaded -------?!?!?
Density Code: 0x58
BlockSize: 0
DataCompEnabled: yes
DataCompCapable: yes
DataDeCompEnabled: yes
CompType: 0x1
DeCompType: 0x1
BOP: yes
Block Position: 0
Partition 0 Remaining Kbytes: 1470031
Partition 0 Size in Kbytes: 1470031
ActivePartition: 0
EarlyWarningSize: 0
NumPartitions: 0
MaxPartitions: 1
"Medium type - not loaded", there was a tape in the drive,
so not sure what that means.
And the resource definitions from the conf files:
------------resource def. from
bacula-dir-------------------------------------------------
# Definition of LTO-5 tape device
Storage {
Name = LTO-5
##Do not use "localhost" here
Address = godwit.domus-bka.local # N.B. Use
a fully qualified name here
SDPort = 9103
Password = "godwitBackup" # password for Storage
daemon
Device = LTO-5 # must be same as
Device in Storage daemon
Media Type = LTO-5 # must be same as
MediaType in Storage daemon
Maximum Concurrent Jobs = 10
}
------------resource def. from
bacula-sd-------------------------------------------------
# A Linux or Solaris LTO-5 tape drive
Device {
Name = LTO-5
Media Type = LTO-5
Archive Device = /dev/st0
AutomaticMount = yes; # when device
opened, read it
AlwaysOpen = yes;
RemovableMedia = yes;
RandomAccess = no;
AutoChanger = no
Maximum File Size = 25GB
## Changer Command = "/usr/libexec/bacula/mtx-changer %c
%o %S %a %d"
## Changer Device = /dev/sg0
## AutoChanger = yes
# # Enable the Alert command only if you have the mtx
package loaded
Alert Command = "sh -c 'tapeinfo -f %c |grep
TapeAlert|cat'"
## If you have smartctl, enable this, it has more info
than tapeinfo
## Alert Command = "sh -c 'smartctl -H -l error %c'"
}
------------end-------------------------------------------------------------------------
Any insight as to what may be happening here, or config
items I've left out would be greatly appreciated!
Thanks much!
|
