A little birdy told me that Anand Avati said: ] GlusterFS puts the best effort at being non blocking. But with recent ] discussions we may make this optional at the protocol level. i've come to understand that, and think the idea of adding it in the future may be helpful (both for the issue i illustrated and those i've read about, like crash recovery)... i was just curious if there were, in particular, some parameters (like timeout values and the like) which might be helpful in introducing an "artificial latency" into the request completion time for the mount process... it seemed a quick (if not completely "clean") solution for this singular issue while still working with the 1.4.0 branch of GlusterFS... ] > Problem #2: this one is strange... i have noticed that OpenOffice ] > (versions 2.0.4 and 2.3.0, at least, on CentOS 4 and 5 respectively) ] > fail to open files on GlusterFS mounts... instead returning the ] > following error: ] > ] > General input/output error while accessing <filename> ] ] Is there anything in the logs of glusterfs at the time of this error? i typically run GlusterFS with loglevel set to ERROR... and with this setting there are no messages during the transaction in either client or server logs on either host... setting the loglevel to DEBUG, i get the following messages in the server log on the appropriate host: 2009-01-17 16:41:36 D [server-protocol.c:3288:server_lookup_resume] home: LOOKUP '13074433/CSE486-HW2.doc' 2009-01-17 16:41:36 D [inode.c:299:__inode_passivate] home/inode: passivating inode(13074839) lru=212/1024 active=6 purge=0 2009-01-17 16:41:36 D [inode.c:280:__inode_activate] home/inode: activating inode(13074839), lru=211/1024 active=7 purge=0 2009-01-17 16:41:36 D [server-protocol.c:3688:server_open_resume] home: OPEN '/simon/Lore/CSE486-HW2.doc (13074839)' 2009-01-17 16:41:36 D [server-protocol.c:6260:server_lk] home: LK 'fd=0' 2009-01-17 16:41:36 D [server-protocol.c:4021:server_flush] home: FLUSH 'fd=0' 2009-01-17 16:41:36 D [server-protocol.c:3923:server_release] home: RELEASE 'fd=0' 2009-01-17 16:41:36 D [inode.c:299:__inode_passivate] home/inode: passivating inode(13074839) lru=212/1024 active=6 purge=0 2009-01-17 16:41:37 D [server-protocol.c:3288:server_lookup_resume] home: LOOKUP '1/simon' 2009-01-17 16:41:37 D [server-protocol.c:3288:server_lookup_resume] home: LOOKUP '12926987/Lore' 2009-01-17 16:41:37 D [inode.c:455:__inode_create] home/inode: create inode(0) 2009-01-17 16:41:37 D [inode.c:280:__inode_activate] home/inode: activating inode(0), lru=212/1024 active=7 purge=0 2009-01-17 16:41:37 D [server-protocol.c:3288:server_lookup_resume] home: LOOKUP '13074433/paperconf' 2009-01-17 16:41:37 D [inode.c:323:__inode_retire] home/inode: retiring inode(0) lru=212/1024 active=6 purge=1 and... all messages logged were indeed DEBUG, and nothing indicates (to me, at least) a problem of any sort... and, of course, all these attempts resulted in the same error message i indicated before... B. Karhan simon at pop.psu.edu PRI/SSRI Unix Administrator
