Yeah, I can see that the cloning would definitely go faster with less
multi-plexing that was done on the source tape, but I'm curious if it
might be possible to multi-plex the savesets to the clone volume just as
they were multi-plexed to the source tape. You could certainly have one
clone operation running against one saveset on one source tape and
another clone operation running against another saveset if you launched
two separate nsrclone operations, so perhaps in the amount of time it
would take you to clone say 25 savesets from one tape you could also
finish 25 from another, but unless you can multi-plex it, it just seems
that doing even 25 would take a conserably long time since it can't do
more than one at at time. I'm assuming it can't, but maybe I'm wrong.
You could always create a tar archive of the data and that could be a
non-NetWorker version, but I'm not aware of any utility that would read
a NetWorker volume and then create a non-Legato output tape. I'm
curious, though, about this business of cloning an entire volume versus
cloning all the savesets on the volume one at a time. They say that when
you're cloning the entire volume, you're really cloning the savesets
themselves. As such, it would appear that cloning a volume would be
tantamount to individually cloning all the savesets one at a time. Has
anyone actually tested this. I wonder if cloning the volume might still
be faster. What they need is a tape copy utility that doesn't pay
attention to multi-plexing and cloning each darn saveset. Just copy the
tape verbatim like dd. Maybe dd could be used to dump the contents of
the tape and then pipe it to tar and then dd it to another tape? Some
such thing?
George
Dave Mussulman wrote:
>
> On Thu, Apr 17, 2003 at 11:11:30AM -0400, George Sinclair wrote:
> > You make an excellent point! I guess I can see that cloning would be
> > reasonably efficient if you were dealing with incrementals or small
> > savesets, but when we run fulls, we're talking about a lot of savesets,
> > some rather large, for a total of maybe 200 savesets or more across as
> > many as 6 tapes. Now, how on earth would I have time to wait for
> > NetWorker to read through each of those savesets and clone each one?
> > Granted, you're not dealing with the network, so you've eliminated that
> > bottleneck, but still, I mean in a 4 drive box, Networker could only
> > clone savesets from two different tapes at a time, which would still
> > leave two more to go. In our other library, it could only clone savesets
> > from one tape at a time since that library only has two drives, which
> > would leave one tape to go.
>
> I'm new to cloning, but I'll interject a point or two here. I haven't
> seen my clones run more than one source and one destination tape, even
> when I feed nsrclone a large list of ssids that I know exist on
> different tapes. Maybe that list needs to be split and delivered to two
> different nsrclone processes running at the same time to utilize writing
> to more than one drive?
>
> Also, I found it very useful to split the clients I was going to clone
> into their own group and run them at a time when other savegroups aren't
> running. This had the advantage of easier-generating the list of ssids
> (now I only had to key off group=ToClone in the mminfo report instead of
> lots of client=XXX fields,) and keeping the multiplexing down (or at
> least limited to a set number of tapes) when they were backing up in the
> first place. That made the cloning go faster.
>
> Cloning is kind of neat, but I'm still not entirely sure it's required
> in my environment. A non-Networker tape-duplicator to store the
> originals offsite would probably work just fine for me. Maybe that
> would work better in your situation as well.
>
> Dave
>
> --
> Note: To sign off this list, send a "signoff networker" command via email
> to listserv AT listmail.temple DOT edu or visit the list's Web site at
> http://listmail.temple.edu/archives/networker.html where you can
> also view and post messages to the list.
> =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
