home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.sgi
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!jvnc.net!yale.edu!ira.uka.de!News.BelWue.DE!news.uni-ulm.de!huober_r
- From: huober_r@hal.rz.uni-ulm.de (Reiner Huober)
- Subject: Re: NFS Question...
- Message-ID: <1992Sep1.135212.11858@wega.rz.uni-ulm.de>
- Sender: huober_r@hal (Reiner Huober)
- Organization: Universitaet Ulm
- References: <1992Aug31.165654.24006@news.iastate.edu>
- Date: Tue, 1 Sep 1992 13:52:12 GMT
- Lines: 31
-
- In 1992Aug31.165654.24006@news.iastate.edu you write
-
- > If a line in my fstab looks like:
- >
- > [machinename]:/usr/mp1 /usr/mp1 nfs bg,soft,timeo=28,retrans=16 0 0
- >
- > should the 'bg' and 'soft' options prevent network lockups when one
- > machine (nfs server) goes down?
-
- generally it's not wise to use the option "soft" on a writeable
- Filesystem (see fstab(4)), since the remote filesystem could become
- corrupted. Also, you must distinguish between "mount options"
- like bg, which are only valid when a fs is mounted, so this will not
- help much when dealing with network lockups, and "connection options".
- We use
-
- ro,bg,soft for read-only mounts (e.g. man pages)
- rw,bg,hard,intr for writeable filesystems (e.g. homedirs)
-
- Network lockups are a common problem using NFS.
- Perhaps you should simply increase the number of nfsd-Daemons
- (Default 4) and biod-Daemons (also 4). Under IRIX3.x, you must edit
- /etc/init.d/network. On our machines (not all SGI's), we use 8 biod
- and 8 nfsd, if a machine exports a filesystem, and on one server, I
- increased the number of nfsd to 16. On our configuration, this led
- to much better behaviour if an important machine goes down, but it did
- not disappear fully :-(
-
- Reiner Huober
-
-
-