Greetings! We're observing a very strange (?) behavior when doing some simple NFS tests: As soon as we add the nfs mount option 'noac' or 'actimeo=0', NFS performance dramatically drops from say 100 MB/s to 2.5 MB/s. I searched archives a little bit and read that this option might have a "slight performance impact", but what we observe is far away from being "slight"... What we tested: Fresh NFS server using Ubuntu 14.04 LTS Fresh NFS server using RedHat Enterprise 6.4 Ubuntu 12.04 client Redhat Enterprise 6.4 Client Oracle 5.7 Client Content of /etc/exports: /shared *(rw,async,no_wdelay,insecure_locks,no_root_squash,no_subtree_check) Mount options: mount -t nfs -o actimeo=0,tcp,vers=3 nfsserver:/shared /mnt/ The mounted /mnt directory only contains a single file. 'Tests' are done using simple "dd if=/mnt/largefile of=/dev/null" Output of 'nfsstat -c': Client rpc stats: calls retrans authrefrsh 1171668 0 1171668 Client nfs v3: null getattr setattr lookup access readlink 0 0% 1159569 98% 0 0% 1 0% 17 0% 0 0% read write create mkdir symlink mknod 12055 1% 0 0% 0 0% 0 0% 0 0% 0 0% remove rmdir rename link readdir readdirplus 0 0% 0 0% 0 0% 0 0% 0 0% 2 0% fsstat fsinfo pathconf commit 0 0% 6 0% 3 0% 0 0% Output of 'nfsstat -s' Server rpc stats: calls badcalls badclnt badauth xdrcall 1265267 0 0 0 0 Server nfs v3: null getattr setattr lookup access readlink 8 0% 1251701 98% 0 0% 1 0% 23 0% 0 0% read write create mkdir symlink mknod 13510 1% 0 0% 0 0% 0 0% 0 0% 0 0% remove rmdir rename link readdir readdirplus 0 0% 0 0% 0 0% 0 0% 0 0% 3 0% fsstat fsinfo pathconf commit 0 0% 8 0% 3 0% 0 0% Any explanation for this? What did I miss? what am I doing wrong? I wouldn't use this option but unfortunately its required for Oracle RAC over NFS... Thanks for any help Markus -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
