Atlas UK

RAL Castor downtime 30/3/11

2011-03-29T10:55:00.002+01:00

Tomorrow, Castor at RAL will be upgraded from 2.1.9 to 2.1.10 to bring it inline with CERN. The downtime is scheduled from 9am till 5pm and during that time, jobs will not be able to run at RAL and files will not be accessible with dq2-get. No new jobs will be allowed to start at RAL after 5pm this evening.

LFC problems over night

2011-03-18T10:57:00.002+00:00

There was a problem overnight at the RAL Tier1 caused by the failure of a network switch stack. This affected a
mixture of disk servers and other grid services. The problem started at around 00:30 local time (=UTC) this
morning Friday 18th March and the network issue was resolved around 03:30 by on-call staff.

Some systems (two out of three lfc front ends and two out of four FTS front end machines) appeared problematic
after the network had been resolved. These systems have now been checked this morning and are running OK.

Because of these problems if you were running grid jobs in the UK overnight, they may have failed when trying to download their input files or register their output on the grid. If you try again now they should work.

ATLAS UK computing tutorial

2011-03-11T17:12:00.002+00:00

An ATLAS UK computing tutorial will be held at NeSC in Edinburgh on Monday/Tuesday 21/22nd March. More information available at http://www.atlas.ac.uk/tutorial2011/

UK cloud offline - LFC intervention

2011-02-08T14:16:00.002+00:00

Tomorrow (Weds 9th Feb.) the UK cloud will be offline for LFC intervention. Without the LFC your jobs don't know where files are stored and you can't register any new data.

This evening (Tues. 8th Feb.) the UK cloud will be set to offline. This will mean that jobs that are already queued will still run but nothing new will be assigned. In general 95%+ of jobs will be drained out of the system in the 12-15 hours before the downtime. At 10:30 UK time the LFC will then be taken down and won't be available again until around 4pm UK time, (assuming things go well). Any jobs that are still running when the LFC goes down will fail as the data won't be registered on the grid. The UK cloud can be considered completely unavailable during this time. Once the LFC is back you will be able to dq2-get your files again. The cloud will be set back online shortly afterwards and job submission will start again. Note some Tier 2 sites will take advantage of the downtime to make upgrades of their own so if you have data at a specific Tier 2 you might want to check it doesn't have a separate downtime.

Registration open for 2011 ATLAS UK meeting

2010-10-29T14:57:00.002+01:00

Registration is now open for the ATLAS UK annual physics meeting, which will take place at RHUL 5-7 January 2011. To find out more about the meeting and to register, please visit http://indico.cern.ch/conferenceDisplay.py?confId=108609. Registration will close at the end of Friday 19 November.

Data distribution within the UK

2010-06-24T13:29:00.003+01:00

This information is for ATLAS UK sysadmins.

ATLAS space tokens and data placement shares have been reviewed for all UK sites. The minimum token size required for ATLAS DATADISK and MCDISK tokens is now 50TB compared to the previous threshold of 25TB. The increased threshold is to alleviate pressure on ATLAS operations when dealing with many small sites.

Some smaller sites are now below the threshold and their share has been reduced to 0%. Sites can increase their share by deploying more disk or re-deploying space tokens. eg. re-deploy DATADISK into MCDISK. Sites should email atlas-support-cloud-uk@cern.ch for advice on how to deploy new disk and how to tweak current token sizes.

This will not affect the number of jobs running at each site as DATA and MC are equally popular. Existing data will remain on all sites until cleaned by ATLAS deletion policy. Regular revision of data share will reflect deployed disk and we expect most sites to increase their storage volume over the coming months, in line with GridPP pledges. A summary of UK site storage and pledge status can be found here.

http://atladcops.cern.ch:8000/drmon/crmon_TiersInfo.html
http://atladcops.cern.ch:8000/drmon/mcmon_TiersInfo.html

Scheduled AT-RISK for LANCS space tokens

2010-06-21T14:51:00.005+01:00

A scheduled downtime affecting Lancaster's disk storage will occur next Monday the 28th of June in order to relocate disk servers to a new data centre. Users trying to access data on ATLAS space tokens will be affected. Site queues will be offline so jobs scheduled to run at Lancaster will be delayed until the downtime has finished.

QMUL in downtime over the weekend

2010-06-17T17:32:00.002+01:00

Queen Mary will go into a downtime tomorrow so that electrical work can be carried out in the building housing the cluster. The downtime is scheduled to last from 10am on Friday 18th June until 6pm on Tuesday the 22nd June.

FTS upgrade will affect file transfers from RAL 17 March 2010

2010-03-16T16:42:00.001+00:00

An upgrade of the FTS service at the RAL Tier1 will happen tomorrow, Wednesday 17th March.

The upgrade is scheduled to take place between 08:00 and 13:00 (local time = UTC) during which time the service will be unavailable. The first hour of the scheduled outage is to drain current transfers. Any remaining queued transfers should start once the upgrade is complete.

New ATLAS collaboration webpages

2010-02-01T14:57:00.001+00:00

The new ATLAS collaboration webpages are now online. http://atlas.web.cern.ch/Atlas/Collaboration/. Follow the ATLAS Live link to reach important recent news.

Pacman mirror to be switched off

2010-01-27T14:39:00.001+00:00

The Tier 1 at RAL currently hosts a Pacman mirror of the ATLAS software. The Pacman mirror is designed for people to use if they are installing their own version of the ATLAS software on their local machine. However the software can also be downloaded direct from CERN which is what most people (who want their own copy of the software) appear to be doing as the RAL mirror has had very low usage in the last year (in December there were just 67 files accessed). Software releases for Tier 2s are automatically downloaded from CERN as well. The machine the RAL mirror is on is being retired soon and unless anybody has any major objections the mirror will not be moved onto a new machine. This machine will be switched off at the start of next week, (1/2/10).

RAL LFC to be off - UK Cloud will be off with it

2010-01-18T16:49:00.002+00:00

The ATLAS UK cloud relies on the Local File Catalog at RAL to find existing files and register new ones. The LFC needs to be taken down to perform essential changes to make it more robust ahead of LHC data taking and the reprocessing beforehand.

As a consequence, the LFC will down from 08:00 to 19:00 on Wednesday (27th); many operations will not be possible in the UK cloud during this period.

We apologize for this disruption. There are hopes that we can be less dependent on the LFC on future, but this requires development work by ATLAS.

Datasets to the UK

2009-12-16T11:09:00.001+00:00

The UK data placement group is beginning to receive requests for datasets to be copied to the Tier 2s UK, sometimes to specific sites. It is perhaps timely to remind people of the computing model constraints on what we do and the ATLAS rules for transfers.

The Tier 2 disk is for Grid access only, and is in principle for all of ATLAS. We consulted the ATLAS community in the UK via the institute physics contacts about specific interests, and we managed to have all of those requests met by having a physics and/or performance group in the UK cloud. ATLAS now believes we have enough groups assigned in the UK. It should not be your expectation that everything you want to access will be in the UK (and certainly not at a geographically local site, if you have one), and there are multiple copies of group areas worldwide. That said, we try to have as much as we can in the UK.

There is also some space for longer-term (but not indefinite) storage of UK user data in a special space token. Again, this would not normally be at a ‘local’ site – this space is for all ATLAS UK users, and so could and should be at any of our sites, which are required to meet minimum standards for data access. This space will be managed by the UK, and unreasonable usage will not be welcomed, just as with maxi disks in the past. ‘The management’ reserve the right to remove data from this space. This space has to hold user-derived datasets, so open-ended subscriptions to a stream or data type cannot go here.

All access to the Tier 2 and collocated UK Grid space must be by the Grid mechanisms. If you have files that you want to access locally, these are by ATLAS policy copied into the scratch area, and you should copy them to some (Tier 3) storage private to your institute/HEPGroup, or for a while and only in modest volume to the UK shared space.

In general people requesting transfers to the UK should be telling us

What need for the data is (activity, which physics and performance group)
Why it is not already covered by a physics or performance group area
How big it will be *at a maximum*
How the data will be used (what sort of job to be run, database access etc)

Clearing of the Tier 2 Data

2009-12-03T11:43:00.002+00:00

SO within the UK it is time for an early christmas spring clean for data at Tier2s.
With this in minf we aare going to be gettting the Tier 2 to remove data from non space token areas. this will end up with extra space being allocated into space tokens fo rusers to use properly.

WE WILL BE REMOVING ALL FILES THAT DO NOT HAVE IN THE SURL OF THE FILE ONE OF THE FOLLOWING FRAGMENTS.
atlasdatadisk
atlasmcdisk
atlasscratchdisk
atlasgroupdisk
atlaslocalgroupdisk
atlashotdisk
atlasproddisk

If all UK users could check to see that all their files ( and particluar their old files) that they want to keep are not going to be deleted.

UK collaboration 2010 - register now

2009-10-29T10:07:00.002+00:00

Registration will close 9th November for the 2010 ATLAS UK collaboration meeting. This will be held in Cambridge, 6-8 January 2010. Register here: http://www.hep.phy.cam.ac.uk/atlasUK2010/register.html

User Analysis Test

2009-10-06T07:42:00.004+01:00

UK grid sites and users should be aware of preparations for a coordinated User Analysis Test (twiki link) towards the end of October. The idea is to further test computing infrastructure under heavy user analysis load, as expected when first data arrives. Please see Jim's talk at the following link: http://indico.cern.ch/conferenceDisplay.py?confId=69437

ATLAS UK collaboration meeting 2010

2009-09-29T16:01:00.002+01:00

Registration is now open for the 2010 ATLAS UK collaboration meeting. This will be held in Cambridge, 6-8 January 2010.

RAL dashboard

2009-09-09T12:46:00.001+01:00

As mentioned at GridPP23 collaboration meeting, RAL has put a Tier1 site status page online: http://www.gridpp.rl.ac.uk/status

ATLAS LFC at RAL changing

2009-08-19T15:42:00.003+01:00

ATLAS will get it's own instance of the UK LFC tomorrow (Thursday 20 August). The new LFC is called lfc-atlas.gridpp.rl.ac.uk so any User Interface machines or user scripts need to be updated if hardcoded.

Notice of dataset cleanup

2009-07-28T11:33:00.002+01:00

Note: datasets remaining in CERN-PROD_USERDISK and CERN-PROD_USERTAPE will be deleted by beginning of September. If this is a problem, please email atlas-uk-comp-users@cern.ch

Distributed computing shifts - a tutorial

2009-07-22T11:18:00.004+01:00

A reminder for UK people interested in contributing to ATLAS Distributed Computing shifts (ADCoS). There is an online tutorial tomorrow (Thursday 23rd) starting at 10am UK time. The tutorial is online via EVO. This is a good opportunity for graduates to contribute to shift work, please ask people to attend. Details are here: http://indico.cern.ch/conferenceDisplay.py?confId=64465

WLCG STEP09 post-mortem

2009-07-08T14:29:00.002+01:00

The WLCG STEP09 post-mortem workshop takes place this Thursday and Friday, agenda and access details at the following link: http://indico.cern.ch/conferenceOtherViews.py?view=standard&confId=56580

LFC back online at RAL

2009-06-26T11:58:00.003+01:00

The UK file catalogue is back online. See the Tier-1 blog for details: http://www.gridpp.rl.ac.uk/blog/2009/06/26/r89-migration-big-friday-26-june/

LFC downtime at RAL

2009-06-25T10:00:00.006+01:00

The UK File Catalogue is in scheduled downtime today (Thurs 25 June), the service is being relocated to a new building at RAL and should be back up later today or early tomorrow. Downtime link: https://goc.gridops.org/downtime/list?id=36505383 and further details about the RAL Tier-1 building move are here: http://www.gridpp.rl.ac.uk/blog/

Decommission of old Manchester storage

2009-06-22T12:15:00.004+01:00

Like many sites in the UK, Manchester now runs the Disk Pool Manager SRM and will decommission the legacy dCache service on July 4th (dcache01 and dcache02.tier2.hep.manchester.ac.uk). If you have private data on these machines, please take action to move the data. This should not affect many users. Here is a link to the announcement.