Re: Lockups!
Posted in 1994
Informix Sysadmin (informix@asl3.asl-labs.bc.ca) wrote: : On Wed, 26 Oct 1994, David Coburn wrote: : > What is in the log when this happens? Anything? : Nothing...just the fact that it has been more than the normal time span : since the last checkpoint. <snip> : The big problem is that when our database goes down, we have 100 : employees who can't do their job. It is a 2 minute fix to bring it down : and up again (albeit VERY ANNOYING), while trying to get support folks : can often be somewhat SLOW. Brian, Next time, do a tbstat -o FILENAME where FILENAME is any name in a directory with enough disk space to contain your entire shared memory plus a little extra. Also do a full "ps" output to a file. Then you can debug to your heart's content against the FILENAME (just like a core dump) and still have your users back up. You say you've checked your tbstat -l. How about the online.log? Any mention of long transactions? What about tbstat -u? If you're 5.0X or greater, the second section of tbstat -u will tell you about any transactions in progress. If you'll post a tbstat -a or email it to me I'll look at it for you. Joe -- --------------------------------------------------------------------------- Joe Lumbley Dallas, Texas Hi, y'all Watch for my INFORMIX DBA Survival Guide from Prentice-Hall/Informix Press ISBN 0-13-124314-4 (12431-3) Due in bookstores October 28, 1994 orders: orders@prenhall.com (515) 284-6761 fax (515) 284-2607 ----------------------------------------------------------------------------