Zabbix for very large environments

[fmrapid]

The Zabbix feature page proclaims support for up to 1 000 000 data points.

Has anyone successfully run an installation over 100K or 200K or 500K. I am evaluating the feasibility of an installation for up to 100K data points before scaling to 500K.

Can it scale to this level or is this reserved for big league commercial real-time data and event management systems. (ex. OSI Soft PI)

[QuakENT]

Yes and no. If you have enough powerfull HW for run such a monitoring environment you can possibly monitor it. On other side... zabbix is still very young project with few developers compare to another even open source solution. So if you will want some advanced function(and you will definitely need them for managing such big thing) to reflect your situation you find, that you must doit yourself. You can browse thru this forum and see, that there is many request for some basic functions but no developer resources for do it, so many despert users doit themselves. So many large installation are very unique in terms of nonofficial features.
I mention that, because if you run this sw especialy on such big environment, you will need a full time dedicated technician(s)&developer(s) for sure. There is some thread, where are users with large installation...just ask them, how much resources they must put into this(such on any other installation of big open source sw projects).