 |
 |
Hi Wilko, yes, the files were written via xrootd. Most likely different servers out of the four servers in the GPFS cluster were used to write the different files. As you know, in the old days the PFNs were representations of the LFNs with a simple substitution. Now they are different and it seems they are not "unique enough" - or we are doing something wrong. Cheers, Andreas On 01/23/2014 07:48 PM, Wilko Kroeger wrote: > Hello Andreas > > How were the files put into xrootd? Were they written through xrootd > (with xrdcp for example) or staged from a remote src ? > If there were written to xrootd were different xrootd servers used ? > > Cheers, > wilko > > > On Thu, 23 Jan 2014, Andreas Petzold wrote: > >> HI, >> >> we are running xrootd 3.3.2 and have configured a bunch of filesystems where xrootd can put the data (actually 4 servers all see the same GPFS filesystems): >> >> oss.space public /export/gka8301/xrootd/ >> oss.space public /export/gka8302/xrootd/ >> oss.space public /export/gka8303/xrootd/ >> oss.space public /export/gka8304/xrootd/ >> oss.space public /export/gka8305/xrootd/ >> oss.space public /export/gka8306/xrootd/ >> oss.space public /export/gka8307/xrootd/ >> oss.space public /export/gka8308/xrootd/ >> oss.space public /export/gka8309/xrootd/ >> oss.space public /export/gka8310/xrootd/ >> oss.space public /export/gka8311/xrootd/ >> oss.space public /export/gka8312/xrootd/ >> oss.space public /export/gka8313/xrootd/ >> oss.space public /export/gka8314/xrootd/ >> oss.space public /export/gka8315/xrootd/ >> >> oss.localroot /export/xrootd/ >> >> I've just discovered files where something seems to have gone very wrong when they were written: >> >> ll /export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c >> lrwxrwxrwx 1 xrootd xrootd 53 Jan 23 10:15 /export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c -> /export/gka8305/xrootd/public/1B/4DA0CE526BED0000176% >> >> ll /export/xrootd/04/31086/82859ca4-8407-11e3-85ef-5b68ff9582bb >> lrwxrwxrwx 1 xrootd xrootd 53 Jan 23 09:22 /export/xrootd/04/31086/82859ca4-8407-11e3-85ef-5b68ff9582bb -> /export/gka8305/xrootd/public/1B/4DA0CE526BED0000176% >> >> getfattr -d /export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c >> getfattr: Removing leading '/' from absolute path names >> # file: export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c >> user.XrdFrm.Pfn="/export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c\000" >> >> getfattr -d /export/gka8305/xrootd/public/1B/4DA0CE526BED0000176% >> getfattr: Removing leading '/' from absolute path names >> # file: export/gka8305/xrootd/public/1B/4DA0CE526BED0000176% >> user.XrdFrm.Pfn="/export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c\000" >> >> getfattr -d /export/xrootd/04/31086/82859ca4-8407-11e3-85ef-5b68ff9582bb >> getfattr: Removing leading '/' from absolute path names >> # file: export/xrootd/04/31086/82859ca4-8407-11e3-85ef-5b68ff9582bb >> user.XrdFrm.Pfn="/export/xrootd/15/65005/e4ab025a-840e-11e3-9317-b38c94feed3c\000" >> >> So there are two different namespace entries pointing to the same file. How can something like this happen? >> >> Cheers, >> >> Andreas >> >> >> ######################################################################## >> Use REPLY-ALL to reply to list >> >> To unsubscribe from the XROOTD-L list, click the following link: >> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1 >> > > ######################################################################## > Use REPLY-ALL to reply to list > > To unsubscribe from the XROOTD-L list, click the following link: > https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1 > -- Karlsruhe Institute of Technology (KIT) Steinbuch Centre for Computing (SCC) Andreas Petzold Hermann-von-Helmholtz-Platz 1, Building 449, Room 202 D-76344 Eggenstein-Leopoldshafen Tel: +49 721 608 24916 Fax: +49 721 608 24972 Email: [log in to unmask] www.scc.kit.edu KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the XROOTD-L list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1