Re: system hangs

In article <72i0fv$plt$1@news.xmission.com>, Nick Nobbe <nnob@loc.gov> writes > >Ken, Cindy, > > I notice this from time to time on my taughtly stretched SCO Unix >3.2.4.2 server, which is running two Online 7.10 instances - one for real, >one for development, with a total of 5 databases (one SE). It's a server >with 64M Ram, running other Unix applications, plus TCP/IP.. > > There are NO messages, as you say, and NO problems, it is just >impossible to start up any more user connections, say, onmonitor, ontape, >dbaccess, and though no more memory has been allocated by IDS, everything >just hangs ... including user sessions. The only cure is to stop >everything, shut down the instance having the problem, and restart. Looking >at CPU usage in onstat and Unix sar, I see a lot of time being spent in >user CPU and very low CPU idle. My guess: I need more of everything. And I >suppose I'll also hear the familiar words, "upgrade", and agree. > > Let me know if you discover anything else that may be causing the problem. > Check you have enough semaphores. I have seen this on a Pyramid machine running an early v7.xx Online. The oninit process had 100% CPU time and even pressing return on the consol did nothing! A pyramid engineer was called who broke out to the boot prom and grovelled in the kernel until he found out what the problem was.. >Yours, > Nick > > > >At 09:23 AM 11/13/98 -0400, Ken Ketchum wrote: >>T> > >>> >System Specs: >>> > Sun Enterprise 5000 >>> > 1 Gb RAM >>> > IDS 7.24.UC5 >>> > >>> >Earlier this week, our database engine hung. There was no error >>> >message, >>> >users could not connect, and we could not shut the database down. We >>> >finally did a 'kill -9' on the processes, rebooted the server, and then >>> >did >>> >an index check. We have no idea what is causing this problem. Informix >>> >support hasn't been able to give us any good clues. >>> > >>> >Anyone have any ideas? >>> > >>> >Thanks, >>> > >>> >cindy >>> > >>> > >>> >========================== >>> >cindy_alford@ymd.yamaha-motor.nl >>> >Yamaha Motor Distribution >> >> >>Are you using all or mostly shared-memory access; and had the >>system been up for a long time? We had very similar behavior >>from 7.24UC5 on a Sparc 1000E runing Solaris 2.5.1. Tech support >>mentioned a bug that wiggled under heavy shared memory use. >>We are working around it by using TCP connections, until we get >>a chance to install 7.30, which supposedly fixes it. >> >>-Ken Ketchum >>Etak, Inc >> >********************* >Nick Nobbe >NLS/BPH >Library of Congress >nnob@loc.gov -- David Williams Maintainer of the Informix FAQ Primary site (Beta Version) http://www.smooth1.demon.co.uk Official site http://www.iiug.org/techinfo/faq/faq_top.html I see you standin', Standin' on your own, It's such a lonely place for you, For you to be If you need a shoulder, Or if you need a friend, I'll be here standing, Until the bitter end... So don't chastise me Or think I, I mean you harm... All I ever wanted Was for you To know that I care