Leslie,
Can you try the attached patch?
Amanda already have a lot of timeout, see 'man amanda-changers'
But Amanda immediately fail on a 'SCSI error; Sense Key=Aborted
Command', try the attached patch.
Jean-Louis
On 04/01/2014 12:43 PM, leslie wrote:
Hello,
Have been using amanda for approximately 3 years and am delighted with
the software.
I have a problem which I believe is related to device sharing.
The OS is CentOS 5.10 and I am running amanda 3.3.3.
Have a SUN SL500 with three LT05 drives, a 90 tape library, and a 15
slot cap loader.
Attached files are failed_12a.rpt, works_12a.rpt, the atape.sh script
(used to manipulate the tape robot), and the amanda.conf file.
Situation:
The AMADSet12a dump process was running and was writing to the
8th tape of the set.
I was using the atape.sh script to empty and load the library for
the next dump set.
The dump failed, this has happened several times in the past, and I
believe it is due to amanda not dealing with a busy tape robot.
Is there a parameter for setting the time that Amanda should
wait for
a tape to load or for the robot to be available to service a
request?
Thank you in advance...have a nice day!
diff Amanda/Changer/robot.pm Amanda/Changer/robot.pm
index 97a7c98..1b363ce 100644
--- Amanda/Changer/robot.pm
+++ Amanda/Changer/robot.pm
@@ -2368,7 +2368,8 @@ sub status {
$err = "SCSI error; Sense Key=$sensekey" if $sensekey;
$counter--;
if (($sensekey eq "Not Ready" and $counter > 0) ||
- ($sensekey eq "No Sense" and $counter > 0)) {
+ ($sensekey eq "No Sense" and $counter > 0) ||
+ ($sensekey eq "Aborted Command" and $counter > 0)) {
return Amanda::MainLoop::call_after(1000, $run_mtx);
}
return $status_cb->("error from mtx: " . $err, {});
