Re: Long checkpoints with no obvious cause

> >Does anything stand out like a sore thumb? > > Not to me. Nothing changed? > >I zero the states each weekend with onstat -z > > Hmmm... I think you might want to call tech support about this, but I seem to remember that we had some problems with onstat -z, and we found out from informix that they didn't recommend using it because there was a bug involved that could crash the instance. I asked when they were planning on releasing a patch, and they said they couldn't because the nature of the problem required more work than just a patch. That was on 7.24, I'm sure, but we upgraded to 7.3? later and I don't _THINK_ it was resolved yet. -- # unrm / ksh: unrm: not found # man cpio Sent via Deja.com http://www.deja.com/ Before you buy.

mars1972@my-deja.com wrote: > > > >Does anything stand out like a sore thumb? > > > > > Not to me. Nothing changed? > > > >I zero the states each weekend with onstat -z > > > > > Hmmm... I think you might want to call tech support about this, but I > seem to remember that we had some problems with onstat -z, and we found > out from informix that they didn't recommend using it because there was > a bug involved that could crash the instance. I asked when they were > planning on releasing a patch, and they said they couldn't because the > nature of the problem required more work than just a patch. > > That was on 7.24, I'm sure, but we upgraded to 7.3? later and I don't > _THINK_ it was resolved yet. That was a nasty problem with an early 7.24 release if memory serves. If it was the problem I remember, it occurred when one added a -r flag, as when running a differential report (such as onstat -p -r 60 -z to get clean stats every minute). Caused a really hectic morning once when my assistent was doing performance monitoring like that and I did not know. The server crashed, I brought it up again. Meanwhile he switched to running his stats on that server's sister which promptly crashed as well so he went back to the now ready first server, and so on. This went on for an hour with neither server up for more than a few minutes at a time and me on the phone yelling at tech support about what happens if this bug infects my other 15 servers? Finally we figured it out. PHEW! Anyway, I believe this one is gone. Art S. Kagel & Family kagel@erols.com