Hi,
at RHUL, Grigori and I are working on the 2.4.0 installation at the
moment. CE, SE, MON and WN have been installed and configured, so far so
good. We are just doing final tests. If we don't uncover any really tricky
problems then we are on course to meet the deadline.
The reason it has taken this long is a combination of limited manpower and
the release being late to start with. We have to fit in the time to work
on this inbetween our other priorities. Hopefully this will be less of a
problem when the RHUL/Brunel London Tier 2 support post is filled.
RHUL is one site which has had a huge amount of scheduled downtime this
year. I think it's well known that this was caused by a security incident.
It was prolonged because, due to limited manpower, we decided in March to
wait for 2.4.0 rather than install 2.3.0, with the prospect of upgrading
in a few weeks' time.
Cheers,
Simon
On Sun, 24 Apr 2005, Coles, J (Jeremy) wrote:
> Dear All
>
>
>
> With only another 3-4 days left to meet the requested upgrade schedule
> I'm pleased to say that many sites are now at 2.4.0. This is excellent
> progress and hopefully a sign of what can be expected with future
> releases. As usual I would urge you to feed back comments on how the
> distribution/process might be improved to your Tier-2 coordinator. For
> those sites that have not yet upgraded please could you give me details
> about the plan and reasons for the delay - I need to feed this back into
> the weekly EGEE operations meeting.
>
>
>
> In addition to looking at upgrade times we have also started looking at
> Site Functional Test histories for sites. Already it is easy to see a
> range of performance. The best performing site appears to be Sheffield
> so well done to the team there - Alessandra, Andrew, Paul and Matthew.
> Unfortunately there are also clear worst performing sites that have
> barely contributed any resources to date and the GridPP Tier-2 board
> will need to consider ways to improve such sites in the near future.
> There are also some sites that have a huge amount of scheduled
> maintenance which needs to be investigated; a work method like this will
> cause problems if it continues once LCG starts running so we need to
> find solutions now to remove the underlying problems.
>
>
>
> Most of you will be fully aware that this week sees the next HEPSYSMAN
> meeting at RAL 26th and 27th:
> http://hepwww.rl.ac.uk/sysman/april2005/agenda.html. There will be
> several talks relevant to maintaining a grid site and working as part of
> the EGEE structure so I encourage everyone who can to attend.
>
>
>
> Many thanks once again to all those people involved in the 2.4.0 upgrade
> it is going very well this time thanks to your efforts.
>
>
>
> Kind regards,
>
> Jeremy
>
>
>
> ------------------------------------------------------------------------
> --------------------------------
>
> Dr Jeremy Coles
>
> GridPP Production Manager
>
> CCLRC eScience Department Phone: (+44)(0)1235 778256
>
>
> Rutherford Appleton Laboratory Fax: (+44)(0)1235
> 446626
>
> Chilton, Didcot, Oxon, OX11 0QX, UK Email: [log in to unmask]
>
> ------------------------------------------------------------------------
> ----------------------------------
>
>
>
>
|