There are some known issues with read-ahead and booster. please try without booster. moreover, booster is effective when loaded on the server side. avati 2008/5/16 nicolas prochazka <prochazka.nicolas@xxxxxxxxx>: > Hi, > I do some test with booster translator, and i ve found some issue : > my test is very simple, client1 -> server1 > > if i do : > > revelation vdisk # dd if=base of=dest bs=10MB > 629+1 records in > 629+1 records out > 6290378752 bytes (6.3 GB) copied, 135.844 s, 46.3 MB/s > > all are ok. I'm not using booster in this case. > > if i do : > > LD_PRELOAD=/usr/local/lib64/glusterfs/glusterfs-booster.so dd if=base > of=test bs=10MB > dd: writing `test': Transport endpoint is not connected > 1+0 records in > 0+0 records out > 0 bytes (0 B) copied, 0.113639 s, 0.0 kB/s > then i do a ls : > revelation vdisk # ls > ls: cannot open directory .: Transport endpoint is not connected > if i do a cd / and then cd /mnt/vdisk , it works again. > > > I'm using big file, sometime, /mnt/vdisk seems to be inacessible, dd > shows this issues all the times. > > > Regards > Nicolas Prochazka > > > > > Version : GlusterFS 1.3.9 > > Client conf file : > volume client1 > type protocol/client > option transport-type tcp/client > option remote-host 10.98.98.1 > option remote-subvolume brick > end-volume > > volume booster > type performance/booster > subvolumes client1 > end-volume > > volume readahead > type performance/read-ahead > option page-size 1MB > option page-count 64 > subvolumes booster > end-volume > > volume iothreads > type performance/io-threads > option thread-count 20 > subvolumes readahead > end-volume > > volume io-cache > type performance/io-cache > option cache-size 1000MB # default is 32MB > option page-size 1MB #128KB is default option > option force-revalidate-timeout 1200 # default is 1 > subvolumes iothreads > end-volume > > volume writebehind > type performance/write-behind > option aggregate-size 512KB # default is 0bytes > option flush-behind on # default is 'off' > subvolumes io-cache > end-volume > > > > > Server conf file > volume brick1 > type storage/posix > option directory /mnt/disks/export > end-volume > > > volume brick > type performance/io-threads > option thread-count 32 > option cache-size 512MB > subvolumes brick1 > end-volume > > > volume readahead-brick > type performance/read-ahead > option page-size 1M > option page-count 128 > subvolumes brick > end-volume > > volume server > option window-size 2097152 > type protocol/server > subvolumes readahead-brick > option transport-type tcp/server # For TCP/IP transport > option client-volume-filename /etc/glusterfs/glusterfs-client.vol > option auth.ip.brick.allow * > end-volume > > > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxx > http://lists.nongnu.org/mailman/listinfo/gluster-devel > -- If I traveled to the end of the rainbow As Dame Fortune did intend, Murphy would be there to tell me The pot's at the other end.
