 |
 |
Hi all, we recently upgraded our XRootD (plus DFS) and it seems that recalling files from tape does not work as expected anymore. I’m now trying to figure out where we messed up. In specific, I am trying to understand how a redirector + 2 staging-server reacts when a client asks for an unstaged file. - I would assumes that if a client requests to open a file, 1) the redirector selects a staging-server which brings the file online, 2) the client gets stalled until the file is online, and 3) the client opens the file when it is online. - Instead, it seems [1] that 1) the redirector stalls the client and asks the staging-servers whether they have the file, 2) the staging-server check their *local* file system and do not find the file, 3) the redirector tells the client that the file does not exist. Do I have to add something to the configuration to go from the second to the first behaviour? Cheers, Max [1] grep ea32d722-0f0f-11e2-9717-5a23a10abeef /var/log/xrootd/manager/* /var/log/xrootd/manager/cmsd.log:191009 11:39:55 15331 SelDFS seeking /12/23095/ea32d722-0f0f-11e2-9717-5a23a10abeef /var/log/xrootd/manager/xrootd.log:191009 11:39:55 19484 jalien.41420:[log in to unmask] ofs_open: 0-600 fn=/12/23095/ea32d722-0f0f-11e2-9717-5a23a10abeef /var/log/xrootd/manager/xrootd.log:191009 11:39:55 15369 Decode f01-124-183-e delays jalien.41420:[log in to unmask] 1 /12/23095/ea32d722-0f0f-11e2-9717-5a23a10abeef /var/log/xrootd/manager/xrootd.log:191009 11:39:56 19484 jalien.41420:[log in to unmask] ofs_open: 0-600 fn=/12/23095/ea32d722-0f0f-11e2-9717-5a23a10abeef /var/log/xrootd/manager/xrootd.log:191009 11:39:56 15369 Decode f01-124-183-e gave jalien.41420:[log in to unmask] err -2 'Unable to access file; file does not exist.' /12/23095/ea32d722-0f0f-11e2-9717-5a23a10abeef ######################################################################## 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