JiscMail Logo
Email discussion lists for the UK Education and Research communities

Help for GRIDPP-STORAGE Archives


GRIDPP-STORAGE Archives

GRIDPP-STORAGE Archives


GRIDPP-STORAGE@JISCMAIL.AC.UK


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

GRIDPP-STORAGE Home

GRIDPP-STORAGE Home

GRIDPP-STORAGE  February 2016

GRIDPP-STORAGE February 2016

Options

Subscribe or Unsubscribe

Subscribe or Unsubscribe

Log In

Log In

Get Password

Get Password

Subject:

xrootd issues at QM

From:

Daniel Traynor <[log in to unmask]>

Reply-To:

Daniel Traynor <[log in to unmask]>

Date:

Mon, 29 Feb 2016 14:21:49 +0000

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (51 lines)

Hi all

This email sent to both atlas and storage groups for info.

I noticed quite a lot of "Get error: adler32 mismatch on input file" errors over the weekend (17/18 Frb) for Atlas analysis jobs running at QMUL, e.g. PandaID: 2776583111, which was some what concerning.

If I understand the pilot job stdout (snippets below) for this job, despite the data file being at QMUL, the job does not use direct access and instead copies the data file in to the scratch directory of the job on the worker node using xrootd (which we can all agree is very stupid  thing to do at a Lustre site). It then does a check of the adler32 checksum and finds this is different from what it expects. 

I've checked and the adler32 checksum on Lustre and it is correct (files are not corrupt at source and are usable in root).

Further investigation (a local user accessing data via xrootd between the grid and HEP clusters) has shown that when our xrootd server is under heavy load then data accessed through it is being corrupted. Checks of the md5sum checksums have shown mismatches between the file on lustre and that provided by xrootd and the md5sum changes each time the file is download via xrootd (when the server is under high load).  

At present the xrootd server is a VM running xrootd version 4.2.3.1 from the xrootd-stable repo (e.g. from http://xrootd.org/ ). I will move it onto physical hardware to see if this helps mitigate the problem. 

Clearly though xrootd should not act in this way. It might be some complicated interaction with Lustre or a bug with xrootd. Are we aware of any similar issues with xrootd? Wheres the best place to put a ticket in for xrootd.

Also the Atlas user should be using direct access in the first place.

cheers

dan
...
28 Feb 17:52:57| FAXSiteMover| SURL=srm://se03.esc.qmul.ac.uk:8444/srm/managerv2?SFN=/atlas/atlasdatadisk/rucio/mc12_8TeV/b4/10/NTUP_COMMON.01414454._026727.root.1
28 Feb 17:52:57| FAXTools.py | sourceSite is not set, use faxredirector value from AGIS
28 Feb 17:52:57| FAXTools.py | Updated redirector for missing protocol: root://atlas-xrd-uk.cern.ch:1094
28 Feb 17:52:57| FAXTools.py | Updated redirector for missing trailing /: root://atlas-xrd-uk.cern.ch:1094/
28 Feb 17:52:57| FAXSiteMover| Will use global path: root://atlas-xrd-uk.cern.ch:1094//atlas/rucio/mc12_8TeV:
28 Feb 17:52:57| FAXSiteMover| GPFN=root://atlas-xrd-uk.cern.ch:1094//atlas/rucio/mc12_8TeV:NTUP_COMMON.01414454._026727.root.1
28 Feb 17:52:57| FAXSiteMover| WARNING: Global LFN not the same as the initial LFN. Will use the global LFN
28 Feb 17:52:57| SiteInformat| Copy tool will be used for stage-in
28 Feb 17:52:57| SiteInformat| not directIn
....
28 Feb 17:52:58| FAXSiteMover| StageIn files started.
28 Feb 17:52:58| FAXSiteMover| Executing command: source /cvmfs/atlas.cern.ch/repo/sw/local/xrootdsetup.sh; export X509_USER_PROXY=/mnt/data/scratch/lcg/pilatl08/home_cream_733838221/cream_733838221.proxy; xrdcp -np root://atlas-xrd-uk.cern.ch:1094//atlas/rucio/mc12_8TeV:NTUP_COMMON.01414454._026727.root.1 /scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1
28 Feb 17:52:58| FAXSiteMover| Using time-out 5071 s for file size 1908523041
28 Feb 17:53:23| FAXSiteMover| Command finished after 24.340000 s: 
28 Feb 17:53:33| FAXSiteMover| Stagein succeeded
28 Feb 17:53:33| FAXSiteMover| Verify file Staging: source: root://atlas-xrd-uk.cern.ch:1094//atlas/rucio/mc12_8TeV:NTUP_COMMON.01414454._026727.root.1, sourceSize: 1908523041, sourceChecksum: af112c68, destFile: /scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1
28 Feb 17:53:33| FAXSiteMover| Getting destination file(/scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1) information.
28 Feb 17:53:33| FAXSiteMover| Getting local File(/scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1) info.
28 Feb 17:53:33| FAXSiteMover| Executing getsize() for file: /scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1
28 Feb 17:53:33| FAXSiteMover| Got file size: 1908523041
28 Feb 17:53:33| FAXSiteMover| Executing adler32() for file: /scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1
28 Feb 17:53:39| FAXSiteMover| Got adler32 checksum: 4291530b
28 Feb 17:53:49| FAXSiteMover| Destination file information: file: /scratch/tmp/9378611.1.sl6_lcg_1G_analysis/condorg_fjdkGKIN/pilot3/Panda_Pilot_25075_1456681422/PandaJob_2776583111_1456681432/NTUP_COMMON.01414454._026727.root.1, size: 1908523041, checksum: 4291530b
28 Feb 17:53:49| FAXSiteMover| !!WARNING!!2990!! Remote and local checksums (of type adler32) do not match for mc12_8TeV:NTUP_COMMON.01414454._026727.root.1 (4291530b != af112c68)



* Dr Daniel Traynor, Grid cluster system manager
* Tel +44(0)20 7882 6560, Particle Physics,QMUL

Top of Message | Previous Page | Permalink

JiscMail Tools


RSS Feeds and Sharing


Advanced Options


Archives

October 2018
September 2018
August 2018
July 2018
June 2018
May 2018
April 2018
March 2018
February 2018
January 2018
December 2017
November 2017
October 2017
September 2017
August 2017
July 2017
June 2017
May 2017
April 2017
March 2017
February 2017
January 2017
December 2016
November 2016
October 2016
September 2016
August 2016
July 2016
June 2016
May 2016
April 2016
March 2016
February 2016
January 2016
December 2015
November 2015
October 2015
September 2015
August 2015
July 2015
June 2015
May 2015
April 2015
March 2015
February 2015
January 2015
December 2014
November 2014
October 2014
September 2014
August 2014
July 2014
June 2014
May 2014
April 2014
March 2014
February 2014
January 2014
December 2013
November 2013
October 2013
September 2013
August 2013
July 2013
June 2013
May 2013
April 2013
March 2013
February 2013
January 2013
December 2012
November 2012
October 2012
September 2012
August 2012
July 2012
June 2012
May 2012
April 2012
March 2012
February 2012
January 2012
December 2011
November 2011
October 2011
September 2011
August 2011
July 2011
June 2011
May 2011
April 2011
March 2011
February 2011
January 2011
December 2010
November 2010
October 2010
September 2010
August 2010
July 2010
June 2010
May 2010
April 2010
March 2010
February 2010
January 2010
December 2009
November 2009
October 2009
September 2009
August 2009
July 2009
June 2009
May 2009
April 2009
March 2009
February 2009
January 2009
December 2008
November 2008
October 2008
September 2008
August 2008
July 2008
June 2008
May 2008
April 2008
March 2008
February 2008
January 2008
December 2007
November 2007
October 2007
September 2007
August 2007
July 2007
June 2007
May 2007
April 2007
March 2007
February 2007
January 2007
December 2006
November 2006
October 2006
September 2006
August 2006
July 2006
June 2006
May 2006
April 2006
March 2006
February 2006
January 2006
December 2005
November 2005
October 2005
September 2005
August 2005
July 2005
June 2005
May 2005
April 2005
March 2005
February 2005
January 2005
December 2004
November 2004
October 2004
September 2004
August 2004
July 2004
June 2004


JiscMail is a Jisc service.

View our service policies at https://www.jiscmail.ac.uk/policyandsecurity/ and Jisc's privacy policy at https://www.jisc.ac.uk/website/privacy-notice

Secured by F-Secure Anti-Virus CataList Email List Search Powered by the LISTSERV Email List Manager