Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uwm.edu!bionet!agate!ucbvax!van-bc!ubc-cs!unixg.ubc.ca!chilko.ucs.ubc.ca!stlouis
From: stlouis@unixg.ubc.ca (Phill St. Louis)
Newsgroups: comp.unix.sysv386,comp.database
Subject: ISC SWAP QUESTION
Message-ID: <1991Apr11.014614.11146@unixg.ubc.ca>
Date: 11 Apr 91 01:46:14 GMT
Sender: news@unixg.ubc.ca (Usenet News Maintenance)
Organization: University of British Columbia, Vancouver, B.C., Canada
Lines: 23
Nntp-Posting-Host: chilko.ucs.ubc.ca

I would like to know what those Interactive users with TCP/IP and
NFS are using for their swap partitions.

What happens when an Interactive Unix system does not have enough
swap space?   Will there be PANICS?  Will processes mysteriously
die?  We are running Oracle and each of the four Oracle processes 
are taking turns issuing the message "process terminated with error".

We are using Interactive UNIX 2.2, running TCP/IP, NFS, ORACLE,
a 386 machine at 33MHz and 12MB of RAM.  The system was configured
with only 11.68MB of swap.  

I would recommend 20MB of swap for this system.  I would like to
know what other people are using for their swap partition and 
what their configuration is. 

Has anyone else had this problem with Oracle processes?  Am I
wrong to say that this is a swap problem?

Any feedback would be much appreciated.

Phill St. Louis
University of British Columbia