Re: 2xRe: urgent : inknown problem

---- you wrote: > Oh yes it is! Didn't I warn this list repeatedly about not running > onstat -g sql or querying sqx_sqlstatement on a busy system with multiple CPUs?> Several times! I am new to the list, so missed that. Is this on any platform/version? ---------------------------------------------------------------- Get your free email from AltaVista at http://altavista.iname.com

All that I'm aware of. Tried Solaris, DGUX on m88k and intel myself. Whether this is a 'bug' of a 'problem' and therefore insoluble is an open question since Informix will not or cannot test it in an SMP production environment. The question comes down to whether the problem is that you can never catch the head of the list because several CPU VPs are moving it along and your sqlexec thread is only one processor and must relinquish the VP periodically or that the tail catches and laps you, for the same reason, and poof you are back behind the head of the list again with the tail chasing you and perhaps your thread is now hopelessly confused. I don't know and neither does Informix. To know if this will affect you ask: Do I have multiple CPUs and multiple CPU VPs? No? Then you are safe. Yes? then YMMV and you will have to try it since it will depend on system load. Art S. Kagel a_blonde@mindless.com wrote: > > ---- you wrote: > > Oh yes it is! Didn't I warn this list repeatedly about not running > > onstat -g sql or querying sqx_sqlstatement on a busy system with multiple CPUs?> > Several times! > > I am new to the list, so missed that. Is this on any platform/version? > > ---------------------------------------------------------------- > Get your free email from AltaVista at http://altavista.iname.com