Emanuele,

Ooops.  Looks like we transposed the names we intended to use in the
original questionnaire response.  And I'd forgotten that was what we
intended to do.

As it happens, we seem to have arrived at a situation where the site CE+GIIS
is also configured as the regional MDS - lcgce01.gridpp.rl.ac.uk - which
matches your `list' of nodes and makes lcgcs01 redundent.  The regional MDS
aspect of lcgce01 isn't configured correctly but is has all the software
bits.

If the load on the MDS is not great, we could operate like that, otherwise
we will use lcgcs01 as the regional MDS box and reconfigure things.  We will
sort it out Monday.

I got no response to my request for write acess to the CVS system (a while
back, now) - perhaps I didn't do it correctly.

Martin.
--
  -------------------------------------------------------
    Martin Bly  |  +44 1235 446981  |  [log in to unmask]
       Systems Admin, Tier 1/A Service,  RAL PPD CSG
  -------------------------------------------------------

> -----Original Message-----
> From: Emanuele LEONARDI [mailto:[log in to unmask]]
> Sent: Friday, August 08, 2003 5:03 PM
> To: [log in to unmask]
> Subject: Re: RAL LCG-1 status
>
>
> Hi Martin.
>
> I see that on lcgce01 you are running the CE and the Site GIIS for
> rallcg1, as also shown on your status page. In your answer to the
> questionnaire, though, this node was tagged as the Regional
> MDS for the
> west region, while lcgcs01 was supposed to be the CE+siteGIIS node.
>
> Are you still planning to run a MDS node to act as Region GIIS for the
> LCG West region? If yes, what node will you be using?
>
> In any case, to enable your site I need to know to which Regional GIIS
> you pointed lcgce01 to. Can you please send me this
> information or, even
> better, commit the current configuration files to CVS?
>
> Thanks
>
>                   Emanuele
>
>
>
> "Bly, MJ (Martin)" wrote:
> >
> > Emanuele,
> >
> > Reinstalled MySQL - still the same RB problem.
> >
> > We've also got the BDII fully configured and it is reporting:
> >
> > ** ... problem adding Mds-Software-deployment=MDS,
> >
> Mds-Host-hn=adc0015.cern.ch,Mds-Vo-name=cernlcg1,Mds-Vo-name=l
> cgeast,Mds-Vo-
> > name=local,o=grid
> > **    Err: 21  mds-service-executable-pid: value #0 invalid
> per syntax
> > server:
> > mds-service-executable-pid: value #0 invalid per syntax
> > ** ... problem adding Mds-Software-deployment=MDS,
> >
> Mds-Host-hn=adc0018.cern.ch,Mds-Vo-name=cernlcg1,Mds-Vo-name=l
> cgeast,Mds-Vo-
> > name=local,o=grid
> > **    Err: 21  mds-service-executable-pid: value #0 invalid
> per syntax
> > server:
> > mds-service-executable-pid: value #0 invalid per syntax
> > ** ... problem adding Mds-Software-deployment=MDS,
> >
> Mds-Host-hn=adc0021.cern.ch,Mds-Vo-name=cernlcg1,Mds-Vo-name=l
> cgeast,Mds-Vo-
> > name=local,o=grid
> > **    Err: 21  mds-service-executable-pid: value #0 invalid
> per syntax
> > server:
> > mds-service-executable-pid: value #0 invalid per syntax
> > ** ... problem adding Mds-Software-deployment=MDS,
> >
> Mds-Host-hn=wn-02-29-a.cr.cnaf.infn.it,Mds-Vo-name=cnaflcg1,Md
> s-Vo-name=lcge
> > ast,Mds-Vo-name=local,o=grid
> > **    Err: 21  mds-service-executable-pid: value #0 invalid
> per syntax
> > server:
> > mds-service-executable-pid: value #0 invalid per syntax
> > ...
> > and so on for all the CEs.
> >
> > every few minutes.
> >
> > It doesn't show our CE in its list (not surprising).
> >
> > Martin.
> > --
> >   -------------------------------------------------------
> >     Martin Bly  |  +44 1235 446981  |  [log in to unmask]
> >        Systems Admin, Tier 1/A Service,  RAL PPD CSG
> >   -------------------------------------------------------
> >
> > > -----Original Message-----
> > > From: Emanuele LEONARDI [mailto:[log in to unmask]]
> > > Sent: Friday, August 08, 2003 11:08 AM
> > > To: [log in to unmask]
> > > Subject: Re: RAL LCG-1 status
> > >
> > >
> > > Hi Martin.
> > >
> > > The RB is always a nasty beast.
> > >
> > > According to past experience, the problem with the LB should
> > > be related
> > > to the setup of the MySQL db. You should first check if you
> > > followed all
> > > the instructions in Appendix C of lcg1-notes.txt. Then
> you may try and
> > > remove the MySQL rpms, clean up the data area and
> reinstall the rpms
> > > from scratch, then repeat the configuration procedure.
> > >
> > > About the annoying message, we get the same thing on our
> RB. It should
> > > not have bad side effects on the short term (no /tmp area
> > > clean up takes
> > > place), so I propose we leave the problem alone and check
> > > with the next
> > > release if it is still there.
> > >
> > > I guess I am the one to prod for activating the whole
> > > merry-go-round to
> > > have your site in. Let me know when your MDS is
> > > up-and-running: RAL will
> > > be the first site in the LCG West region, so I have to
> reconfigure my
> > > BDII to read info from it. Once it works for me, I'll
> publish the new
> > > settings so that all BDII's can be reconfigured the same way.
> > >
> > > Please do not forget to commit to CVS the node
> configuration files you
> > > are currently using.
> > >
> > > Cheers
> > >
> > >                   Emanuele
> > >
> > > "Bly, MJ (Martin)" wrote:
> > > >
> > > > Greetings,
> > > >
> > > > The RAL LCG-1 set is up and we are able to do various
> > > things locally which
> > > > indicate at least some of it is working.  We have some
> > > outstanding issues:
> > > >
> > > > 1/ edg-job-list-match works OK and reports the local CE,
> > > but edg-job-submit
> > > > fails:
> > > >
> > > > $ edg-job-submit testJob.jdl
> > > >
> > > > **** Error: API_NATIVE_ERROR ****
> > > > Error while calling the "LOG::regist" native api
> > > > Unable to Register the Job to the LB logger at:
> > > lcgrb01.gridpp.rl.ac.uk:9002
> > > >
> > > > 2/  The RB is kicking out this (email) message with
> > > monotonous regularity...
> > > >
> > > > To: [log in to unmask]
> > > > Subject: Cron <root@lcgrb01>
> /opt/edg/sbin/edg-wl-purgeStorage -l
> > > >     /opt/edg/var/log/edg-wl-purgeStorage.log -t 604800 -a 40
> > > >
> > > > Unable to set sandbox dir from the environment: EDG_WL_TMP
> > > not defined...
> > > >
> > > > Bright ideas at this point would be appreciated -
> > > otherwise, we're about
> > > > ready to join in - if someone can tell me who @ CERN i
> need to prod?
> > > >
> > > > Martin.
> > > > --
> > > >   -------------------------------------------------------
> > > >     Martin Bly  |  +44 1235 446981  |  [log in to unmask]
> > > >        Systems Admin, Tier 1/A Service,  RAL PPD CSG
> > > >   -------------------------------------------------------
> > >
> > > --
> > > /------------------- Emanuele Leonardi -------------------\
> > > | eMail: [log in to unmask] - Tel.: +41-22-7674066 |
> > > |  IT division - Bat.31 2-012 - CERN - CH-1211 Geneva 23  |
> > > \---------------------------------------------------------/
> > >
>
> --
> /------------------- Emanuele Leonardi -------------------\
> | eMail: [log in to unmask] - Tel.: +41-22-7674066 |
> |  IT division - Bat.31 2-012 - CERN - CH-1211 Geneva 23  |
> \---------------------------------------------------------/
>