On Wed, 5 Mar 2008, Kadlecsik Jozsef wrote: > Digging into the mailing list archive brought out some GFS tuning options > too. What is already done: > > - statfs_slots set to 128 > - statfs_fast set to 1 > > What we are going to try is to set > > - glock_purge to 50 > - demote_secs to 150 The new settings did not really help, unfortunately. I believe our performance problem might come from bouncing locks: there is no way to deliver incoming mail systematically to the same machine where the user is actually logged in or connected to via POP/IMAP. So when we see bad performance that may be the result of delivering to one machine and accessing the mailbox from another one. Which knobs should we tune in order to achieve the fastest lock processing? Also, currently we use maildirs. Were mailbox format better suited to GFS, from locking point of view? Best regards, Jozsef -- E-mail : kadlec@xxxxxxxxxxxx, kadlec@xxxxxxxxxxxxxxxxx PGP key: http://www.kfki.hu/~kadlec/pgp_public_key.txt Address: KFKI Research Institute for Particle and Nuclear Physics H-1525 Budapest 114, POB. 49, Hungary -- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
