Re: [Opensim-dev] Memory cache

2009-05-28 Thread diva
Imaze, I'll be looking forward to seeing your contribution coming on mantis. Along with Melanie's improved Cache.cs, and the simplistic GlynnTucker cache, we'll have a good set of mem cache options to compare. One thing I'd like to point out to everyone is that this particular piece of

Re: [Opensim-dev] Memory cache

2009-05-28 Thread Frisby, Adam
...@metaverseink.com Sent: Thursday, 28 May 2009 7:40 AM To: opensim-dev@lists.berlios.de Subject: Re: [Opensim-dev] Memory cache Imaze, I'll be looking forward to seeing your contribution coming on mantis. Along with Melanie's improved Cache.cs, and the simplistic GlynnTucker cache, we'll have a good

Re: [Opensim-dev] Memory cache

2009-05-27 Thread Dahlia Trimble
From a quick glance at the code it looks like it's keeping seldom used objects in memory and increasing memory usage. This may be causing portions of memory to swap out where they may not have before. When retrieving the asset, it would need to swap it back in and it may have to do it in several

Re: [Opensim-dev] Memory cache

2009-05-27 Thread diva
My hunch is the same as yours, i.e. that a file system cache is the right thing to do here. There's no point in holding on to assets in memory after they have been sent to clients, given how fast they can be retrieved from the file system when the next client comes along. Since this is all

Re: [Opensim-dev] Memory cache

2009-05-27 Thread Justin Clark-Casey
d...@metaverseink.com wrote: I noticed we have GlynnTucker.Cache.dll in the distribution, and this wasn't being used. I have no idea what the story is, but since it's there, I did one alternative cache module implementation using it, just to see what happens. Since everything is modularized

Re: [Opensim-dev] Memory cache

2009-05-27 Thread diva
Justin Clark-Casey wrote: fyi, the GlynnTucker library was being used until recently in the original OpenSim/Framework/Communications/Cache/AssetCache.cs, which is why it's there. Ahhh! That explains the mystery. Sorry, it's hard to put all these historical pieces together, I only started

Re: [Opensim-dev] Memory cache

2009-05-27 Thread Melanie
In fact, it must be a BSD compatible license for all parts to be eligible for core. If any part is not licensed under a BSD compatible license, then it would have to live on Forge, which hosts other licenses. Melanie d...@metaverseink.com wrote: Imaze, this is awesome! Thank you so much for

Re: [Opensim-dev] Memory cache

2009-05-27 Thread Imaze Rhiano
Hi Again! Adding BSD compatible licence to the cache (CnmHashGenerationCache) is not problem (if I have understand right BSD licence still allows me to use it my closed source database project). I also have some ideas how to add time based expiration and how user could define to configuration

Re: [Opensim-dev] Memory cache

2009-05-27 Thread Melanie
Hi, Imaze Rhiano wrote: Adding BSD compatible licence to the cache (CnmHashGenerationCache) is not problem (if I have understand right BSD licence still allows me to use it my closed source database project). As the original creator, you can license the code as BSD to OpenSim, but later

Re: [Opensim-dev] Memory cache

2009-05-27 Thread Stefan Andersson
Imaze, if you are doing work from scratch, you always have the option to keep a proprietary version for yourself, whilst releasing one version as BSD. As long as you're not using code that any other OpenSim developer has committed, all rights still belong solely to you. Before jumping

Re: [Opensim-dev] Memory cache

2009-05-26 Thread diva
Just a quick note to inform everyone that word from osgrid folks indicates that the asset cache is still borked and still a mystery. When it is on, CPU usage spikes to unsustainable values (150% with 4 avies in WP). If anyone cares to take a look at it / replace it with something else, please

Re: [Opensim-dev] Memory cache

2009-05-26 Thread Arthur Valadares
If we ran a profiling with this cache maybe we could see where the bottle neck is, couldn't we? Mono has simple statistical profiler that shouldn't be too intruding: $ mono --profile=default:stat program.exe If anyone who saw this performance loss could test this, would be interesting to see

Re: [Opensim-dev] Memory cache

2009-05-26 Thread diva
I noticed we have GlynnTucker.Cache.dll in the distribution, and this wasn't being used. I have no idea what the story is, but since it's there, I did one alternative cache module implementation using it, just to see what happens. Since everything is modularized now, replacing the asset cache

Re: [Opensim-dev] Memory cache

2009-05-21 Thread Stefan Andersson
: a...@deepthink.com.au To: opensim-dev@lists.berlios.de Date: Wed, 20 May 2009 17:21:51 -0400 Subject: Re: [Opensim-dev] Memory cache We are using an external caching library I believe, it could be sitting in there. Adam -Original Message- From: opensim-dev-boun

Re: [Opensim-dev] Memory cache

2009-05-21 Thread Mic Bowman
regards, Stefan Andersson   From: a...@deepthink.com.au   To: opensim-dev@lists.berlios.de   Date: Wed, 20 May 2009 17:21:51 -0400   Subject: Re: [Opensim-dev] Memory cache     We are using an external caching library I believe, it could be sitting in there.     Adam

Re: [Opensim-dev] Memory cache

2009-05-21 Thread Melanie
then be to break the binaries apart into byte cucks smaller than the LOH trshold. Best regards, Stefan Andersson From: a...@deepthink.com.au To: opensim-dev@lists.berlios.de Date: Wed, 20 May 2009 17:21:51 -0400 Subject: Re: [Opensim-dev] Memory cache We are using an external

Re: [Opensim-dev] Memory cache

2009-05-21 Thread Nebadon Izumi
. Best regards, Stefan Andersson From: a...@deepthink.com.au To: opensim-dev@lists.berlios.de Date: Wed, 20 May 2009 17:21:51 -0400 Subject: Re: [Opensim-dev] Memory cache We are using an external caching library I believe, it could be sitting

Re: [Opensim-dev] Memory cache

2009-05-21 Thread Mic Bowman
would then be to break the binaries apart into byte cucks smaller than the LOH trshold. Best regards, Stefan Andersson   From: a...@deepthink.com.au   To: opensim-dev@lists.berlios.de   Date: Wed, 20 May 2009 17:21:51 -0400   Subject: Re: [Opensim-dev] Memory cache     We

Re: [Opensim-dev] Memory cache

2009-05-21 Thread diva
trshold. Best regards, Stefan Andersson From: a...@deepthink.com.au To: opensim-dev@lists.berlios.de Date: Wed, 20 May 2009 17:21:51 -0400 Subject: Re: [Opensim-dev] Memory cache We are using an external caching library I believe, it could

Re: [Opensim-dev] Memory cache

2009-05-21 Thread Melanie
: a...@deepthink.com.au To: opensim-dev@lists.berlios.de Date: Wed, 20 May 2009 17:21:51 -0400 Subject: Re: [Opensim-dev] Memory cache We are using an external caching library I believe, it could be sitting in there. Adam -Original Message- From: opensim-dev-boun

Re: [Opensim-dev] Memory cache

2009-05-21 Thread diva
would then be to break the binaries apart into byte cucks smaller than the LOH trshold. Best regards, Stefan Andersson From: a...@deepthink.com.au To: opensim-dev@lists.berlios.de Date: Wed, 20 May 2009 17:21:51 -0400 Subject: Re: [Opensim-dev] Memory cache We are using

Re: [Opensim-dev] Memory cache

2009-05-21 Thread diva
Mic Bowman wrote: Maybe I'm missing something because I have no idea what I'm right about... Hehehe. You're right in pointing to that piece of code, which is not used anymore, and it's critically important for this puzzle :-) That old code is Framework.Cache.AssetServerBase and

[Opensim-dev] Memory cache

2009-05-20 Thread diva
Here's an unexpected outcome of the wobbling last weekend. Many sims in OSGrid are now experiencing huge performance improvements, both in memory consumption and general speed -- much faster teleports, script loadings, etc. This got me completely puzzled, because we didn't do anything to

Re: [Opensim-dev] Memory cache

2009-05-20 Thread Nebadon Izumi
Yay for mistakes that end in gains!!! of course if anyone asks i planned this ;P anyway i thought people would like to see my ini changes.. this is what lead to the discovery: [OpenSim.ini] ; The following is the configuration section for the new style grid servers ; If you don't know

Re: [Opensim-dev] Memory cache

2009-05-20 Thread Frisby, Adam
@lists.berlios.de Subject: Re: [Opensim-dev] Memory cache The surprise really is that the cache would be a performance loss instead of a performance gain. What is there that makes it so slow? Melanie Frisby, Adam wrote: As chief engineer aboard the USS OSgrid, I might want to recommend against

Re: [Opensim-dev] Memory cache

2009-05-20 Thread Melanie
...@lists.berlios.de] On Behalf Of Melanie Sent: Wednesday, 20 May 2009 1:56 PM To: opensim-dev@lists.berlios.de Subject: Re: [Opensim-dev] Memory cache The surprise really is that the cache would be a performance loss instead of a performance gain. What is there that makes it so slow? Melanie