I have finally been successful in getting an ISM restore working to a
different machine. It has been a tough haul but we got there in the
end. I say we, but to be brutally frank, I got there without any
little bit of help from my friends, neither OpenPSL or IBM.
I did have some support From Robert Reissaus who has been there before.
It's funny but even that kind of support can be of value.
The final solution will need some investigation before I can publish
but if you get sent the 39 steps from Informix (8 on the source and 31
on the target) the problem is step 19 on the target. I redid this step
manually using the nsradmin visual interface and the system came into
life. I did have a few device linking errors and shared memory config
problems (it is Linux after all) but I have seen those before when
trying to use onbar with other storage managers.
Please look out in the near future for "The 39 steps". An Informix
adventure.
regards
Malcolm
R.A. Reissaus wrote:
> Malcolm.
> I know of this problem because we have it as well. So far we haven't been
> able to solve it, and were forced to go back to ontape; wich doesn't present
> any problem.
>
> Would you be willing to send me this alternative set you got from IBM tech?
> I would be most gratefull.
> Regards,
> Robert A. Reissaus
>
> ------------------------------------------------------------------------
> R.A. Reissaus
> Risdi, Amsterdam, the Netherlands
> Ibm/Informix Consultants
> Telefoon: +31(0)20 6331140 : Phone
> Mobiel : +31(0)6 51185352 : Mobile
> E-mail : r.a.reissaus@risdi.com
> Website : www.risdi.com
> ------------------------------------------------------------------------
> -----Oorspronkelijk bericht-----
> Van: owner-informix-list@iiug.org [mailto
wner-informix-list@iiug.org]
> Namens mweallans@panacea.co.uk
> Verzonden: maandag 26 september 2005 15:46
> Aan: informix-list@iiug.org
> Onderwerp: Re: ISM restore to different machine
>
>
> Thanks Andreas,
>
> I think I followed all the steps but it is a very convoluted procedure. It
> also assumes that I have control of the source system. In fact in parallel
> IBM tech support have provided me with a different set of instructions.
> That too specifies what I have to do on the source system (8 steps) and what
> to do on the destination system (31 steps).
>
> I am just about to set out on the "Thirty Nine Steps". (Now what does that
> remind me of?)
>
> BTW I also enpountered a problem in that I didn't see anywhere any warnings
> about running ISM backups as root. But if you don't run ISM backups as root
> the bootstrap does not get copied. And that presents problems later on.
> But when I modified the source system to run all ism backups as root the
> backup times grew from 20 minutes to 11 hours. The reason - PRINTER
> environment variable. It wasn't set but it still seemed to cause a problem.
> (Funnier and funnier)
>
> regards
>
> Malcolm
> (up to my neck in crocodiles - now which swamp was I in when I
> started?)
>
>
> Andreas Breitfeld wrote:
> > Hi,
> >
> > please check if you did all steps described in 'IBM Informix Storage
> > Manager Administrator's Guide' chapter 'Transferring Data with the
> > Imported Restore'.
> >
> > If you get an error like
> > nsrmm: error, file disk data1 not in media index
> >
> > when running the ism_op-mount command on the target machine then run
> > 'ism_catalog -recreate_from' command for each of your devices befor.
> >
> > Andreas
> >
> >
> > On Wednesday 21 September 2005 21:58, malcolm weallans wrote:
> > > IDS 9.40 on Suse Linux SLES8.
> > >
> > > We backup IDS to disk using onbar and ISM. We then backup the whole
> > > Linux disk using cpio. And we have recovered this configuration to
> > > the same physical machine in the past. We are now trying to "clone"
> > > the machine for further development purposes. We restored all of
> > > the informix directories and the ISM backup directories. We then
> > > created the appropriate devices for ISM, and mounted them. So far
> > > so good. We started up ism and asked ism_catalog to find_bootstrap
> > > on the device. It failed to find a bootstrap. It did manage to
> > > recover the save set info though.
> > >
> > > Does anybody have any idea as to the contents of the bootstrap?
> > >
> > > Needless to say that - as usual - the development project haven't
> > > allowed us much time to do this "delicate" operation.
> > >
> > > And BTW - we did have propblems with ism_startup but we solved those
> > > by unsetting the PRINTER environment variable and by installing the
> > > c compiler. It's interesting that ism needs those facilities to run
> > > on Linux. I Live and Learn.
> > >
> > > Malcolm
> > > sending to informix-list
> > sending to informix-list
> sending to informix-list