Paul,
On Wed, Jun 28, 2006 at 10:38:45AM +0200, Paul Bijnens wrote:
> On 2006-06-26 17:29, Brian Cuttler wrote:
> ># df -kh /amanda/work
> >Filesystem size used avail capacity Mounted on
> >/dev/dsk/c7t6d0s0 67G 118M 66G 1% /amanda
> >
> >With a reserve value of 10, we should be ablet o utilize 60 Gig
> >of work area, or do I have some value set backwards ?
>
>
> I believe the "reserve" value is not relevant here.
>
> The reserve value comes into play when there is a tape problem.
> At that moment, Amanda falls back to "plan B": it will dump
> incrementals instead of fulls, now only to holdingdisk. This
> mode is called "degraded mode".
ah... I was over exagerating the importance of "reserve", with no
use of the work area and reserve commented out (default 100 %) I
forgot the important, "reserve is only looked at if the tape is
unavailable" conditional.
There is (fortunately) no such thing as, always reserve in case
the tape fails or fills. I'd been thinking there was but I did
need to be reminded that this isn't the case.
> The default value is to reserve all of the holdingdisk for
> these "degraded" mode backups (incrementals instead of fulls).
> If you want to allow some full backups in this mode too, then you
> should lower the value of reserved space for incrementals.
I have no idea what a proper value is, the data types are too mixed
and there are too many clients for me to have any sort of a feel for
what is right to do at my site. I suppose a higher reserve is "better"
(and achieved by default by leaving reserve commented out) as it allows
the maximum space and give at least an opportunity to get incrementals
of the largest number of DLEs.
> The problem that some DLEs seem to be bypassing the holdingdisk is
> not the result of the "reserve" parameter.
Yes, I see that now. I could use more work area, especially since I
will likely be adding additional large DLE, my (Solaris) based LNotes
server has insufficent tape capacity, that will probably move over as
soon as I commit the C2-Jukebox/LTO3 to service in place of the L9/LTO
jukebox now in use.
I also noticed that I seemed to have some delays on bandwidth yesterday
and increased the threshhold, it may have been a fortunate combination
of levels for by DLEs but we where back to a nice normal 8 hour elapse
run for 27 plus hours of dump time.
I'm going to mark this down as a success.
Thank you Jon, Stefan, Paul.
Brian
STATISTICS:
Total Full Daily
-------- -------- --------
Estimate Time (hrs:min) 0:11
Run Time (hrs:min) 27:26
Dump Time (hrs:min) 8:34 5:41 2:54
Output Size (meg) 167914.8 133469.1 34445.7
Original Size (meg) 167914.8 133469.1 34445.7
Avg Compressed Size (%) -- -- -- (level:#disks ...)
Filesystems Dumped 56 6 50 (1:43 2:6 3:1)
Avg Dump Rate (k/s) 5572.0 6688.7 3383.3
Tape Time (hrs:min) 25:58 20:38 5:20
Tape Size (meg) 167914.9 133469.2 34445.8
Tape Used (%)84.0 66.7 17.3 (level:#disks ...)
Filesystems Taped 56 6 50 (1:43 2:6 3:1)
Avg Tp Write Rate (k/s) 1839.3 1839.6 1838.1
> --
> 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 *
> ***********************************************************************
>
---
Brian R Cuttler brian.cuttler AT wadsworth DOT org
Computer Systems Support (v) 518 486-1697
Wadsworth Center (f) 518 473-6384
NYS Department of Health Help Desk 518 473-0773
|
