Not mad, definitely worth a poke around. I noticed this first with 1.10 on a production site. It would chug away happily during the day then die when I first opened a management screen the next morning. It seemed that it was that first management screen which caused bother - when serving ordinary pages after a long period of inactivity, it seemed happier.... A cron script to restart at 2am every morning has been running for a long time now and it certainly helps.

Dieter Maurer wrote: > > Then I activated a checker, that checks availability > every 5 minutes. The problem went away. Since then, > Zope never hung again. > > Maybe, Zope does not like to be lazy?

Joking aside, I'm almost inclined to agree with this. Our server that has a habit of dying can lie dormant for a few hours and that's when it seems likely to die most.

Am I mad or is this worth looking into?