 |
 |
Hi Elvin, Makes sense to me. I suppose this is only an issue for Castor. In normal xrootd if delete is specified the OFS plugin automatically adds r/w mode (i.e. update). I guess that doesn't occur in the Castor plugin. Andy On Fri, 18 Sep 2015, Elvin Sindrilaru wrote: > Ok, concerning this last issue what do you say if I modify the code not to rely on the user to put also the write/update flag when using the delete one, but do it explicitly in the client code. So this line: > https://github.com/xrootd/xrootd/blob/master/src/XrdCl/XrdClFileStateHandler.cc#L1434 > will change to: > > if (pOpenFlags & kXR_delete) > pOpenFlags &= (~kXR_delete | kXR_open_updt); > > This would avoid the scenario where the client requests an open for write (using only the delete flag) and then in the back the recovery mechanism opens it only in reading mode, and the client tries to do a write and gets an error. Using the extra kXR_open_updt is safe as the user anyway asked for the kXR_delete in the first place. > > > --- > Reply to this email directly or view it on GitHub: > https://github.com/xrootd/xrootd/issues/287#issuecomment-141405144 --- Reply to this email directly or view it on GitHub: https://github.com/xrootd/xrootd/issues/287#issuecomment-141407950 ######################################################################## 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