I wonder if it not only because I use the pool "Default Clone" which I cant
modify and contain as device, all my adv_lun and my tapes.
As the error is waiting for for a volume in tape or disk pool, might be
because, I have all devices in "default Clone" pool.
What would happen if I create a staging pool with only the 2 tapes devices.
My guess, it would open more than one session to my tape !!
Any ideas, if my guess is right ?
Or does someone know for sure, that "It's one session per drive, and that the
end of it"
Regards
Marcel
-----Message d'origine-----
De : Ronny Egner [mailto:RonnyEgner AT gmx DOT de]
Envoyé : mercredi, 28. avril 2010 15:45
À : Hirter Marcel; NETWORKER AT LISTSERV.TEMPLE DOT EDU
Objet : Re: RE: [Networker] staging parralelism
-------- Original-Nachricht --------
> Datum: Wed, 28 Apr 2010 15:12:00 +0200
> Von: Hirter Marcel <Marcel.Hirter AT ne DOT ch>
> An: "NETWORKER AT LISTSERV.TEMPLE DOT EDU" <NETWORKER AT LISTSERV.TEMPLE DOT
> EDU>
> CC: \'Ronny Egner\' <RonnyEgner AT gmx DOT de>
> Betreff: RE: [Networker] staging parralelism
> Thank Ronny,
> Here is what I have:
> I have 5 adv_lunx 1 to 5, and 2 tapes drives.
> Staging policy 80% (high water) - 65% (low water).
> When one adv_lun reach 80% the staging to one tape drive start, when a
> second adv_lun reach 80% the second drive is used, when a third adv_lun reach
> 80%, I got an alert, saying "waiting for an tape or disk for Default Clone
> pool".
> So, as the writing to tape is slower than the writing to disk, I got
> another drive reaching 80%, before one of the two drive are freed.
Having 5 adv_devices with 2 tape makes no sense from my point of view
(because your can have a maximum of zwo parallel staging
processes).
In addition to that at the moment networker chooses devices in order
they were labled; not in order of free space. And if a device fills
up networker insists to write this specific device and does not
continue writing on any other adv_device.
There is a blog post of Preston about this:
http://nsrd.info/blog/2010/04/26/adv_file-operational-changes
Anyway you can used your staging area to keep your data as long as
possible. In my point of view you should keep your data for a specified
amount of days and size your staging area accordingly.
In my environment i keep 8 days of data at the staging area because i do
a full every 7 days... so if i need to recover data i have them on disk
(and cloned on tape anyway) and my staging area is sized accordingly.
To fix your issues you can either buy faster devices (i assume the
bottleneck is your tape and not your disk) or lower the threshold
so the staging has more time to stage before the devices fills up
or wait for the new adv_devices changes.
Yours sincerely
Ronny Egner
--
Ronny Egner
RonnyEgner AT gmx DOT de
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems with this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
|
