On Monday 15 October 2012 16:08:00 Davina Treiber wrote:
> On 15/10/12 12:32, Dag Nygren wrote:
> > Hi all
> >
> > Just investigating why my clone job, which should clone all full savesets
> > that hasn't yet been cloned to tape, but skipped certain savesets and found
> > the following:
> >
> > The Clone resource has very simple settings:
> >
> > - Make only one copy
> > - Filter on Pool: AllPool, Level: full and everything since 6 months ago.
> >
> > The preview list is indeed empty...
> >
> > But the output of the following command, which should be the same thing, is
> > not empty:
> >
> > mminfo -q "copies<2,pool=AllPool,level=full,savetime>6 months ago" -r
> > "client,name,copies,savetime(20),clretent(20),ssid(15),volume,ssflags"
> >
> > The command returns all of those savesets that should have been cloned, but
> > were not by the resource.
> > Dont't like the idea of of the "Clone resource" leaving certain savesets
> > uncloned...
> >
> > Now my questions are:
> >
> > - Is this a known bug?
> > - Does the media database query from the Clone resource include other
> > criteria?
> > - Workaround?
> >
> > The Networker version is 8.0.0.3.
> >
> > Best
> > Dag
> >
> > PS! Tried to enter this on support.emc.com, but got hit with 1 million
> > "Internal server error" and gave up...
>
>
> Have you tried checking the flags of the save sets in your mminfo query?
> You might consider !incomplete, !suspect, !ssercycle.
Hi again!
Your answer made me think more about this. I just requested ssflags and found
nothing strange about those. Doing a strings on nsrtask I noticed that
the part wher nsrclone was unvoked had some "suspect" strings close by and
(re)reading the mminfo
man page I noticed that suspect is not part of ssflags, but clflags. And that
was it! All the savesets not cloned were in
suspect state. (Which is fairly strange on a AFTD...) Clearing the flag made
"NSR clone" pick up the savesets.
Thanks for steering me in the right direction!
Best
Dag
|