Amanda-Users

Amanda -- seems to be running forever

2003-06-13 10:33:20
Subject: Amanda -- seems to be running forever
From: "Michael Packer" <pac AT stingrayboats DOT com>
To: <amanda-users AT amanda DOT org>
Date: Fri, 13 Jun 2003 09:46:33 -0400



> This will allow  to work like amanda wants to workm and you will
> then have a minimum of two generations of everything at full plus any
> incrementals since the last full on everything.
>
> To repeat to the point of boreing the list:
>
> dumpcycle is the number of days or weeks that you give amanda to make
> a full, level 0 backup of everything in your disklist.
>
> runspercycle tells amanda how many times she will be run in
> 'dumpcycle' time.
>
> tapecycle is the number of tapes in the pool for amanda to use.
> Because its considered unwise to attempt to append to an already used
> tape, amanda expects a tape that is either new, or is a 'timed out'
> tape, having outlived its 'tapecycle' time in the pool for each fresh
> run of amdump.

i think that clears it up for me...for some reason i didn't understand it
the 8 other times i read it...i'd suggest they take this wording and put it
out there..it does make a lot more sense..

someone asked about my disklist...

localhost sbd1 always-full
localhost sba2 always-full
localhost sda7 always-full
localhost sda8 always-full
localhost sda5 always-full
localhost sda3 always-full
localhost sda9 always-full

i've included a new copy of the my amanda.conf and am running it now to see
what the system does.

I'm not sure if the DLT1 is correct or not..i couldn't find an exact match
for it on the web in different configurations i read about.

I appreciate the help!!!  Anything I can do to drop ArcServeIT will a major
accomplishment!


(update: using the configuration attached, the back up has been running
since 1:45 am....my load has been as high as 10 [this is a dual 1.1g]  ..
it's now 9:45 and it's still writing to tape)


pac

Attachment: amanda.conf
Description: Binary data

<Prev in Thread] Current Thread [Next in Thread>