home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!pipex!warwick!uknet!sys.uea.ac.uk!gel
- From: gel@sys.uea.ac.uk (Gareth Lee)
- Newsgroups: comp.unix.ultrix
- Subject: Re: Load average way too high
- Message-ID: <gel.727966499@zen.sys.uea.ac.uk>
- Date: 25 Jan 93 12:54:59 GMT
- References: <1993Jan22.145823.10967@julian.uwo.ca>
- Sender: news@sys.uea.ac.uk
- Organization: UEA School of Information Systems, Norwich, UK.
- Lines: 22
- Nntp-Posting-Host: zen.sys.uea.ac.uk
-
- john@phobos.sscl.uwo.ca (John Tucker) writes:
-
- >I hope some one has some clues on this.
-
- >Every once in a while with no real pattern, the load average (as reported
- >by w -u) climbs and gets up around 3.50.
-
- >There are at most 3 or 4 users logged in at any time, so this seems way too
- >high. A 'ps -aux' reveals that 'inetd' and 'syslog' are the major
- >consumers of %CPU and %MEMORY.
-
- I wonder if you have a frequently recurring error (perhaps in inetd)
- which syslog is logging. They should neither require significant CPU
- or memory - my 'ps -aux' entries are (after 2 weeks uptime).
-
- daemon 90 0.0 0.6 164 144 ? I 0:01 /etc/syslog
- root 149 0.0 0.6 184 148 ? I 0:23 /etc/inetd
-
- Try looking in /var/spool/mqueue/syslog for any error messages or
- run 'uerf -R | more' to see if any ultrix error messages are logged.
-
- Gareth Lee.
-