 |
 |
Hi Elvin, This is weird. rc=-252 should never happen. Getting a core file that I can look at next week when this occurs would be a good idea. I'd like to understand why that number is there (well, that assumes that xrootd was compiled on a machine where errno numbers correspond to the machine where it is being executed). Andy On Thu, 17 Sep 2015, Elvin Sindrilaru wrote: > I managed to trigger the StreamTimeout but "unfortunately" the recovery works fine ... So, what I do now artificially is to sleep in the open in the Castor plugin a bit more than the StreamTimeout and what I get on the server side is: > > ... first open request here ... > > Then after the stream timeout: > > 150917 11:26:15 7766 XrdInet: Accepted connection from [log in to unmask] > 150917 11:26:15 7766 XrootdXeq: esindril.729:22@esdss001 pub IPv4 login as esindril > 150917 11:26:15 7766 esindril.729:22@esdss001 XrootdProtocol: endsess 7759:20.3 > 150917 11:26:15 7766 esindril.729:22@esdss001 XrootdProtocol: endsess 7759:20.3 rc=-252 (Unknown error 252) > 150917 11:26:15 time=1442481975.904163 func=open level=INFO logid=253105f6-5d1e-11e5-b5ef-02163e006e2c [log in to unmask]:1094 tid=140144431539968 source=XrdxCastor2Ofs:467 tident=esindril.729:22@esdss001 path=/castor/cern.ch/dev/e/esindril/dir_default/test2.dat, opaque=castor.accessop=3&castor.client_sec_gid=1028&castor.client_sec_uid=58602&castor.exptime=1442482035&castor.id=esindril.729:9@esdss001&castor.manager=lxc2dev6.cern.ch:1094&[log in to unmask]&castor.pfn2=0:15511:1fb4fc92-6bbd-e250-e053-8d06100af1ca&castor.sfn=/castor/cern.ch/dev/e/esindril/dir_default/test2.dat&castor.signature=AAfdRGNe6WKdza6fcCQ7QHORLS+5rwk/lhZHGNXVNcS37ASWheDgOprTe/KfAhdjaiwOJwFAeqjb7zpmJ7ZUww==&castor.txtype=user&oss.asize=1730&tried=lxc2dev6d1.cern.ch,lxc2dev6d1.cern.ch,lxc2dev6d1.cern.ch, isRW=1, open_mode=4200, file_ptr=0x7f75dc002e00 > 150917 11:26:15 7766 esindril.729:22@esdss001 castor2ofs_open: 4200-40644 [log in to unmask] > > But as you can see the rc=-252 from the endsess is different from the rc=4 from the real case. So, I'm still not triggering exactly the same thing. > > Now, when it comes to waitresp, we don't send them explicitly from the Castor OFS plugin, but this only sits on top of the default OFS, so if this returns SFS_STARTED then we also return this. > > > --- > Reply to this email directly or view it on GitHub: > https://github.com/xrootd/xrootd/issues/287#issuecomment-141027853 --- Reply to this email directly or view it on GitHub: https://github.com/xrootd/xrootd/issues/287#issuecomment-141037813 ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the XROOTD-DEV list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1