g00f schrieb:
> Tape Size (meg) 1446225.1 1446225.1 0.0
Here is where it gets fuzzy... and not in logic. Tape size... the
> amount that was put to tape?
Yes.
> Tape Used... 359%?
Habe a look at your tapetype definition. There is a length of the tape
defined, and I'm sure it doesn't be the correct value for your tape. Amanda
writes to the tape until it's full and when the length value is smaller than
the real one, you get unrealistic values here.
http://wiki.zmanda.com/index.php/Tapetype_definitions
# man amtapetype
> Avg Tp Write Rate (k/s) 57752.5 57752.5 --
Avg Tp Write Rate, the holding disk wrote to the tape drive @
> 57MB/s....correct?
Not the holdingdisk writes to tape. Amanda reads from the holding disk and
flushes the data to tape with the shown speed.
How did my Tapes load more then their capacity of 360GB
> each (LTO3 @ 400GB native capacity - 40GB tape split size).
Have a look at the wiki link for tapetype definitions. You can use one of
them, if you find your changer there. Or you can create your own for better
values. If you create a own for your hardware, please share it on the wiki or
send it to the list (someone will add it then).
taper: Will request retry of failed split part.
I cant tell if this is an error or just for my information, I assume
> it caused by the Tape Split Size value we use.
I have no idea on this message. But it seams to be an information. If
something wasn't done successfully, you would see an "FAILED" in the summary like
genome.mr.lfmg.de /tmp lev 0 FAILED "[disk /tmp, all estimate timed out]"
archive2.win -TS/033x010 0 722933590 722933590 -- 198:16 60769.8 241:30
49893.5
archive2.win -S/148Ax050 0 758000870 758000870 -- 346:00 36512.1 185:53
67962.4
And finally this final output which tells me that bandwidth was different for
> the 2 FS's , why would one throughput be higher then the other one.
Maybe there was some load on the server because of different services on the
first filesystem. Or maybe if you have configured to run two dump processes on
the server at the same time, maybe one was finished earlier so the second
could run faster after that time.
Regards
Marc
--
Marc Muehlfeld (Leitung IT)
Zentrum fuer Humangenetik und Laboratoriumsmedizin Dr. Klein und Dr. Rost
Lochhamer Str. 29 - D-82152 Martinsried
Telefon: +49(0)89/895578-0 - Fax: +49(0)89/895578-78
http://www.medizinische-genetik.de
|