[Opensim-dev] Client packet throttling - kilobits versus kilobytes

Frisby, Adam adam at deepthink.com.au
Mon Oct 6 09:30:35 UTC 2008


I'd fix asap. That would explain some issues I've had too.

Adam

> -----Original Message-----
> From: opensim-dev-bounces at lists.berlios.de [mailto:opensim-dev-
> bounces at lists.berlios.de] On Behalf Of Justin Clark-Casey
> Sent: Saturday, 4 October 2008 2:51 PM
> To: opensim-dev at lists.berlios.de
> Subject: [Opensim-dev] Client packet throttling - kilobits versus
> kilobytes
>
> [Sorry, maximum settable in the Linden UI should have read 1500 kbits,
> not 15000 (also corrected below to save confusion)]
>
>
>
> In LLPacketThrottle we interpret the AgentThrottle numbers coming from
> the client as kilobytes per second.  This is also
> detailed in libomv docs at
>
> http://www.libsecondlife.org/docs/trunk/html/T_OpenMetaverse_AgentThrot
> tle.htm
>
> But the Linden Lab client UI lists lists its maximum network traffic
> figure in kilobits per second, where the maximum
> settable is 1500.  At this setting, the
> AgentThrottle packet comes through with a total throttle max of 1536000
> (which does approximate to 1500 kbits).
>
> However, as we're interpreting this as a bytes per second figure
> instead, we end up setting a max throttle of 12000 kbits.
>
> This may be the cause of some of the various 'bomb crater' and 'ghost
> prim' bugs that have crept up recently (notably
> http://opensimulator.org/mantis/view.php?id=2283), where we end up
> pushing data to the client faster than the network
> link can handle it.
>
> I am finding it a little hard to believe that this has gone unnoticed
> for so long though, so I'd like to know what other
> people think about this.
>
> --
> justincc
> Justin Clark-Casey
> http://justincc.wordpress.com
> _______________________________________________
> Opensim-dev mailing list
> Opensim-dev at lists.berlios.de
> https://lists.berlios.de/mailman/listinfo/opensim-dev



More information about the Opensim-dev mailing list