[Opensim-dev] Kudos MW

[Brianna]

I agree with your kudos due.. my observation is my Suse servers are within 80% of my Net servers on a performance measure. How memory leaks, I have yet to determine with mono and the r9950+ 

Huggers to Y'all

  ----- Original Message ----- 
  From: James Stallings II 
  To: opensim-dev at lists.berlios.de 
  Sent: Monday, June 29, 2009 8:04 AM
  Subject: [Opensim-dev] Kudos MW


  Greetings, List,

  Just wanted to take the opportunity to give the OpenSim dev team a robust pat on the back, an attaboy, a wtg peeps, a standing ovation, and most importantly the heartfelt thanks and admiration  of myself and my colleagues on the OSGrid management team, for the work reflected in the last 50 or so svn revisions; not to take anything away from the 9900 or so prior revs *winks* but these last 50 or so include some that have positively impact performance and stability in profound ways.



  Some that come immediately to mind as contributing to the improvement:

  r9951: MW 
  Removed the List<NewForce> m_forceList from ScenePresence, as there wasn't any need for a list, as only the last entry in it was acted on. So it now has a single NewForce m_nextVelocity , which is updated (rather than a
  NewForce object being created every AgentUpdate). So as well as cutting out all the adds and clearing of the list, it also removes the creation of upto 100+ new objects per second per avatar.


  r9934: Dr Scofield for Allan Webb

  After noticing on several occasions that the thread counts
  we see when running OpenSIm were bordering on the astronomical
  I decided to seriously investigate.

  After much poking I discovered that the problem is actually very
  simple. The XEngine secition of the example ini says that the
  timeout for an iden thread is in seconds, and an example value
  of 60 is specified. In fact, this is actually resulting in a 60mS
  idle timeout, which is not normally enough for a smart thread
  to survive. I have added a multiplier to the XEngine constructor
  so that the number now matches the published behavior.


  r9925-9938: Dr Scofield

  MUCH work to streamline and optimize packet-handling code

  r9924: Dr Scofield for Alan Webb 

  Nice patch to streamline and optimize texture processing

  ...and many more :)

  We have seen available memory on our linux plaza boxes go from ~380MB as of ~r9940 to a little over 1GB since updating to r9952. 

  There's still a few problems - Lbsa Plaza was not accepting logins this morning, but it was still running - and that after a dozen hours straight of high-traffic uptime. This is a serious improvement, as we have been seeing sim crashes there on a nearly hourly basis in recent times.

  Thanks Y'all :D

  Keep up the good fight :D

  Cheers!

  James
  aka Hiro_P/daTwitch/Hiro Protagonist

  -- 
  ===================================
  http://osgrid.org
  http://del.icio.us/SPQR
  http://twitter.com/jstallings2
  http://www.linkedin.com/pub/5/770/a49




------------------------------------------------------------------------------


  _______________________________________________
  Opensim-dev mailing list
  Opensim-dev at lists.berlios.de
  https://lists.berlios.de/mailman/listinfo/opensim-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://opensimulator.org/pipermail/opensim-dev/attachments/20090630/b52644aa/attachment-0001.html>