Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uwm.edu!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!midway!ux1.cso.uiuc.edu!news.iastate.edu!sharkey!nstar!jwt!john
From: john@jwt.UUCP (John Temples)
Newsgroups: comp.unix.sysv386
Subject: Re: ESIX FFS: what's the verdict?
Message-ID: <1991Feb8.160031.27585@jwt.UUCP>
Date: 8 Feb 91 16:00:31 GMT
References: <27B16F3A.26B6@tct.uucp>
Organization: Private System -- Orlando, FL
Lines: 21

In article <27B16F3A.26B6@tct.uucp> chip@tct.uucp (Chip Salzenberg) writes:
>Is the FFS reliable enough for a root partition?  How about a /u partition?

I've been running the ESIX FFS on a combined root/usr, a /usr2, and a
/usr/spool/news partition for about 8 months now with no
data-threatening problems.  But I do have a UPS on my system, so it
never goes down unexpectedly.  It does have two problems that I'm
aware of (I'm sure there are more): if you run out of inodes, the
system panics; and you can't successfully [f]stat() a named pipe to
determine the number of bytes in the pipe.  I've reported both to
ESIX and gotten no bug fixes, or promises thereof.

There is also a problem when you're using the BSD mode of the FFS
(which gives you long filenames) in that the system directory reading
routines will not find all of the files in a directory.  I've never
used the BSD mode, but others in this newsgroup have.

Another note: if you use the S51K filesystem, it has the "missing
inode bug."
-- 
John W. Temples -- john@jwt.UUCP (uunet!jwt!john)