Separation of class & sub-class is exactly what I had done and had reduced the Windows class machines to about 40. In addition, I increased the "hubtimeout" from 30 to 120 on almost all the daemons, including that of zenwin and zenmodeler. This is because all daemons kept complaining that zenhub is not available.
For a few hours everything appear to be fine until I added some Windows Services to be monitored and then all daemons and zenwin in particular started to die to the point that I had to write while loop to re-start everything at a minute intervals (don't laugh). The other symptom is the appearance of "bad wmi state" errors and monitoring stopped on all windows machines . WARNING zen.zenwinmodeler: skipping wkssv001.bwc.org has bad wmi state We are at threshold of version 2.2 and yet the software is so buggy that it can not be considered. Not even by a techie who can't afford commercial software. After all, a network monitoring tool is useless, if it can not be trusted. interestingly, even zenping keeps dying! forget about claims of monitoring 5000 nodes.... -------------------- m2f -------------------- Read this topic online here: http://community.zenoss.com/forums/viewtopic.php?p=18740#18740 -------------------- m2f -------------------- _______________________________________________ zenoss-users mailing list [email protected] http://lists.zenoss.org/mailman/listinfo/zenoss-users
