speed, if I use interleaved jobs on disks, copytotape takes very long...

On Wed, Jan 4, 2017 at 9:05 PM, Bruno Friedmann <[email protected]>
wrote:

> On mercredi, 4 janvier 2017 08.35:12 h CET Robert N wrote:
> > On Wednesday, January 4, 2017 at 5:13:35 PM UTC+1, Bruno Friedmann wrote:
> > > On mercredi, 4 janvier 2017 05.30:13 h CET Robert N wrote:
> > > > Hi Bruno,
> > > >
> > > > thanks for your reply.
> > > > I've got it in the end working with Device = "| magic script" I
> think,
> > > > but
> > > > unfortunately doesn;t solve the problem. I think with @| bareos will
> > > > execute it only once at startup, because it's a config include, I
> needed
> > > > it
> > > > executed every time a resource is being used. Nevertheless, the
> problem
> > > > is
> > > > always the same, when bareos has more than 1 device to write to, the
> > > > jobs
> > > > fail with the "No volume name given error", see here :
> > > > https://groups.google.com/forum/?fromgroups=#!topic/
> bareos-users/Nbv-_fR
> > > > Wxa
> > > > Y.
> > > >
> > > > I tried 3 diff scenarios:
> > > >
> > > > 1. define multiple devices in storage resource, as recommended in
> bareos
> > > > doku. 2. script the device definition as described above.
> > > > 3. use a virtual autochanger as described here:
> > > > http://blog.bacula.org/whitepapers/CommunityDiskBackup.pdf
> > > >
> > > > the problem is the volume selection I think, when more than 1 device
> is
> > > > available in the storage.
> > > >
> > > > any clues how to solve this?
> > > >
> > > > thanks,
> > > > Robert
> > >
> > > I'm using mono device for each storage location always since too long.
> > > I've no autochanger and if I have to use one (when doing training) I'm
> > > using mvhtl (available at bareos)
> > >
> > > What I'm doing when I want to merge several backups jobs coming from
> > > several backup clients is using spooling. And this for several years
> now,
> > > has always works, and when the infrastructure permit it, the storage do
> > > all the compression with autoxflate plugin.
> > >
> > > Why, do I do this, because multiple device was initially implemented
> and
> > > made for autoloaders with all the mechanics and so on....
> > > ;-)
> >
> > thanks.
> >
> > what I'd like to achieve is, to run more jobs in parallel, doing 1
> > job/volume. not sure how could I achieve this with spooling and only one
> > device in storage. talking about backup to disk.
>
> I you want (the why is to determine, cause it can be a good idea, as a bad
> one) to achieve one job per volume then you will need one storage per job.
>
> Joba -> poolA -> storage A (with device type A) -> sd / device A
>
> What is the motivation to have this one job per volume when using disk ?
>
>
> --
>
> Bruno Friedmann
>  Ioda-Net Sàrl www.ioda-net.ch
>  Bareos Partner, openSUSE Member, fsfe fellowship
>  GPG KEY : D5C9B751C4653227
>  irc: tigerfoot
>
> openSUSE Tumbleweed
> Linux 4.9.0-2-default x86_64 GNU/Linux, nvidia: 375.26
> Qt: 5.7.1, KDE Frameworks: 5.29.0, Plasma: 5.8.4, kmail2 5.4.0
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "bareos-users" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/
> topic/bareos-users/iZhmf9IfE1c/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> [email protected].
> To post to this group, send email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"bareos-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to