Re: [Opensim-dev] some scalability tests...

2010-01-27 Thread Kyle
Incredible work Mic thanks for helping us understand the most reliable paths to running Opensim considering the may options... From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-boun...@lists.berlios.de] On Behalf Of Mic Bowman Sent: Wednesday, January 27, 2010 2:58 AM To:

Re: [Opensim-dev] some scalability tests...

2010-01-27 Thread diva
More coolness! WRT opening up child agents in neighboring regions. Right now there's a very simple, but too inflexible, rule: the only child agents opened are those in immediate regions, and they open both the UDP and the CAPs connections. The inflexibility comes from our fuzzy understanding

Re: [Opensim-dev] some scalability tests...

2010-01-27 Thread diva
Err.. that's EstablishAgentCommunication, not EnableChildAgent. Same initials, different order :-) d...@metaverseink.com wrote: More coolness! WRT opening up child agents in neighboring regions. Right now there's a very simple, but too inflexible, rule: the only child agents opened are

Re: [Opensim-dev] [Opensim-users] some scalability tests...

2010-01-27 Thread Mic Bowman
Getting the fixed size expanded view isn't too difficult (it was maybe 10 lines of code). Getting it tied to far clip setting in the viewer is a bit more interesting. More because I don't know when the agent update packet comes in with the far clip set relative to the time when the child agents

[Opensim-dev] Region Scalability (was: Re: Hi All)

2010-01-27 Thread Morgaine
Hi Umar. Attempting to increase region scalability by region subdivision is broken by design. We examined many different scaling models over two years in the Architecture Working Group in SL, and region subdivision is the worst proposal by far. I wrote up a fairly detailed analysis of this very