Red Hat Enterprise 3 doesn't seem to have strace as a command.
I thought rather than killing the processes manually, I'd reboot the server
and see if amcleanup runs as included in /etc/rc.d/rc.local (thought I may
as well test that).
Now the server came back up, and none of the amanda services are active
anymore (unsurprisingly). Nothing seemed to happen, so I did a manual
amcleanup, with these results:
amcleanup: no unprocessed logfile to clean up.
Scanning /mnt/hdb1...
20060714: found Amanda directory.
So I'm thinking that this backup run is now finally broken.
Next I thought I'll run amflush and see what happens. It outputs this:
Scanning /mnt/hdb1...
20060714: found Amanda directory.
Today is: 20060717
Flushing dumps in 20060714 to tape drive "/dev/nst0".
Expecting tape daily-1 or a new tape. (The last dumps were to tape daily-3)
Are you sure you want to do this [yN]? y
Running in background, you can log off now.
You'll get mail when amflush is finished.
Now what I notice is that it asks for the tape called daily-1, whereas the
tape I used for Friday's backup was daily-3. Does this mean that daily-3
was filled up and caused this whole issue?
Which brings me to another question. I've used these tapes before for
testing. Will Amanda have appended Friday's backup to what was already on
the tape daily-3, or does it overwrite data previously written to that tape
each time a new backup runs? The reason I ask this is that the tape drive
capacity is 160GB, and I believe that I'm trying to back up a lot less data
than that.
After I rebooted, I got this email from Amanda. As you can see, it only
used 4.7% of the tape:
*** THE DUMPS DID NOT FINISH PROPERLY!
These dumps were to tape daily-3.
The next tape Amanda expects to use is: daily-1.
FAILURE AND STRANGE DUMP SUMMARY:
cerberus /.fonts.cache-1 lev 0 FAILED [disk /.fonts.cache-1 offline on
cerberus?]
cerberus /.autofsck lev 0 FAILED [disk /.autofsck offline on cerberus?]
STATISTICS:
Total Full Daily
-------- -------- --------
Estimate Time (hrs:min) 0:04
Run Time (hrs:min) 0:16
Dump Time (hrs:min) 3:07 3:07 0:00
Output Size (meg) 56785.8 56785.8 0.0
Original Size (meg) 136236.1 136236.1 0.0
Avg Compressed Size (%) 41.7 41.7 --
Filesystems Dumped 107 107 0
Avg Dump Rate (k/s) 5169.8 5169.8 --
Tape Time (hrs:min) 0:13 0:13 0:00
Tape Size (meg) 7259.3 7259.3 0.0
Tape Used (%) 4.7 4.7 0.0
Filesystems Taped 104 104 0
Avg Tp Write Rate (k/s) 9801.6 9801.6 --
USAGE BY TAPE:
Label Time Size % Nb
daily-3 0:13 7259.3 4.7 104
And then, after I ran amflush, I got an email saying this (I didn't actually
put daily-1 into the drive):
*** A TAPE ERROR OCCURRED: [cannot overwrite active tape daily-3].
Some dumps may have been left in the holding disk.
Run amflush again to flush them to tape.
The next tape Amanda expects to use is: daily-1.
And when I now do amstatus daily, I get:
Using /var/lib/amanda/daily/amflush.1 from Mon Jul 17 12:58:42 BST 2006
minerva:/home 0 8774296k waiting to flush
minerva:/usr/local/clients 0 32253287k waiting to flush
minerva:/usr/local/development 0 9687648k waiting to flush
I feel a headache coming on again...
Any suggestions as how to best proceed?
Paul Bijnens wrote:
>
> On 2006-07-17 13:32, Joe Donner (sent by Nabble.com) wrote:
>> and ps -fu amanda outputs:
>>
>> UID PID PPID C STIME TTY TIME CMD
>> amanda 2136 2135 0 Jul14 ? 00:00:00 /bin/sh /usr/sbin/amdump
>> daily
>> amanda 2145 2136 0 Jul14 ? 00:00:02 /usr/lib/amanda/driver
>> daily
>> amanda 2146 2145 0 Jul14 ? 00:00:52 taper daily
>> amanda 2147 2146 0 Jul14 ? 00:00:34 taper daily
>> amanda 2148 2145 0 Jul14 ? 00:12:55 dumper0 daily
>> amanda 2153 2145 0 Jul14 ? 00:00:19 dumper1 daily
>> amanda 2154 2145 0 Jul14 ? 00:00:00 dumper2 daily
>> amanda 2155 2145 0 Jul14 ? 00:00:00 dumper3 daily
>>
>> Does this tell anyone anything?
>
> It means the processes are still alive.
>
> Just a wild guess... Maybe you have specified a manual changer, and
> Amanda is just waiting for you to manually insert the next tape?
>
> Now find out what they are doing, and why it takes days to proceed.
>
> As root or amanda you can trace a process and see if it does somehting
> else, or is just sleeping on some event that will not happen:
>
> strace -p pid-of-the-process
>
> There are two taper processes, one reads from the holdingdisk file
> into a shared memory region, while the other one writes the bytes
> from shared memory to tape. When there is no holdingdisk file, then
> maybe the reader-taper is reading from a network socket?
> And maybe you specified a long dtimeout?
>
>
> --
> Paul Bijnens, xplanation Technology Services Tel +32 16 397.511
> Technologielaan 21 bus 2, B-3001 Leuven, BELGIUM Fax +32 16 397.512
> http://www.xplanation.com/ email: Paul.Bijnens AT xplanation DOT com
> ***********************************************************************
> * I think I've got the hang of it now: exit, ^D, ^C, ^\, ^Z, ^Q, ^^, *
> * F6, quit, ZZ, :q, :q!, M-Z, ^X^C, logoff, logout, close, bye, /bye, *
> * stop, end, F3, ~., ^]c, +++ ATH, disconnect, halt, abort, hangup, *
> * PF4, F20, ^X^X, :D::D, KJOB, F14-f-e, F8-e, kill -1 $$, shutdown, *
> * init 0, kill -9 1, Alt-F4, Ctrl-Alt-Del, AltGr-NumLock, Stop-A, ... *
> * ... "Are you sure?" ... YES ... Phew ... I'm out *
> ***********************************************************************
>
>
>
--
View this message in context:
http://www.nabble.com/Unravel-amstatus-output-tf1953587.html#a5359829
Sent from the Amanda - Users forum at Nabble.com.
|
