paul@eye.com (Paul B. Booth) (02/15/91)
Here's a summary of the responses I got to my posting re. settings for minfree and inode density in hp-ux kernels. Many thanks to those who responded. Thoughts on minfree: The consensus seems to be that 5% minfree could well cause performance problems in the long run on full disks with a lot of activity. Most folks counseled that 10% is more reasonable, but that it might be excessive on larger (>500Mb) disks. In any case, it was brought to my attention that minfree can be adjusted up or down at any time (with tunefs). One interesting observation from Hurf Sheldon: > All said before is correct except when in a heavily NFS > dependent environment. It appears that when the NFS server goes > below 20% free space and there are heirarchical read/writes (someone > compiling in a server directory on a client) performance drops like a stone. > Read only server situations don't exhibit this behavior as > badly but a client generated find on a 90% full NFS disk will go much > slower than on an 80% full disk. (I don't know why - perhaps fragmentation > causing more disk accesses being queued ) Thoughts on inode density: People generally felt that the default inode density (1024 bytes/inode) wastes a lot of space. Generally, folks recommended 8192 bytes/inode, but counseled that it might be best to look at a mature filesystem with 'bsd -i' to get a better feeling for how # of files stacks up against # of free inodes. Point was also made that different filesystems may have different requirements depending on their use. Contributors: rocky@hpfcmgw.HP.COM (Rocky Craig) hurf@batcomputer.tn.cornell.edu (Hurf Sheldon) munir@hpfcmgw.HP.COM (Munir Mallal) Graham Eddy <graham@hparc0.aus.hp.com> "Brian Bartholomew" <bb@reef.cis.ufl.edu> -- Paul B. Booth (paul@eye.com) (...!hplabs!hpfcla!eye!paul) ------------------------------------------------------------------------------- 3D/EYE, Inc., 2359 N. Triphammer Rd., Ithaca, NY 14850 voice: (607)257-1381 fax: (607)257-7335