No subject

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=
 steps in order to locate the asset. This may cause further delay as it may=
 have to swap other memory out to make space. It would seem likely that a f=
ile based cache may perform better as the filename might provide a more dir=
ect route to the actual cache data. I suspect that the operating system may=
 also allow searching directories for a filename fairly quickly - modern fi=
le systems are quite efficient these days.<div>
<br></div><div>This is really just a guess... I suspect some benchmarking a=
nd profiling of several systems of different configurations may be necessar=
y to determine what the bottleneck really is.<br><br><div class=3D"gmail_qu=
ote">
On Tue, May 26, 2009 at 3:52 PM,  <span dir=3D"ltr"><<a href=3D"mailto:d=
iva at metaverseink.com">diva at metaverseink.com</a>></span> wrote:<br><block=
quote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc=
 solid;padding-left:1ex;">
I noticed we have GlynnTucker.Cache.dll in the distribution, and this<br>
wasn't being used. I have no idea what the story is, but since it's=
<br>
there, I did one alternative cache module implementation using it, just<br>
to see what happens. Since everything is modularized now, replacing the<br>
asset cache implementation is as simple as editing your XXXCommon.ini<br>
and changing the name:<br>
<br>
;AssetCaching =3D "CoreAssetCache"<br>
AssetCaching =3D "GlynnTuckerAssetCache"<br>
<br>
I would still urge that someone figures out what's wrong with<br>
OpenSim/Framework/Cache.cs, which is the one used in CoreAssetCache.<br>
<br>
The GlynnTuckerAssetCache has been tried in the UCI Grid and in some<br>
sims in OSGrid and at least it's not being an anti-cache. Let's wai=
t a<br>
few more days to see if it's being a real cache; if so we should see<br=
>
declines on the traffic to the asset servers :-)<br>
<div><div></div><div class=3D"h5"><br>
Arthur Valadares wrote:<br>
> If we ran a profiling with this cache maybe we could see where the<br>
> bottle neck is, couldn't we?<br>
><br>
> Mono has simple statistical profiler that shouldn't be too intrudi=
ng:<br>
><br>
> $ mono --profile=3Ddefault:stat program.exe<br>
><br>
> If anyone who saw this performance loss could test this, would be<br>
> interesting to see where it's spending all this extra time.<br>
><br>
> On Tue, 2009-05-26 at 10:57 -0700, <a href=3D"mailto:diva at metaverseink=
.com">diva at metaverseink.com</a> wrote:<br>
>> Just a quick note to inform everyone that word from osgrid folks<b=
r>
>> indicates that the asset cache is still borked and still a mystery=
. When<br>
>> it is on, CPU usage spikes to unsustainable values (150% with 4 av=
ies in<br>
>> WP).<br>
>><br>
>> If anyone cares to take a look at it / replace it with something e=
lse,<br>
>> please do.<br>
>><br>
>> OpenSim/Framework/Cache.cs<br>
>> _______________________________________________<br>
>> Opensim-dev mailing list<br>
</div></div>>> <a href=3D"mailto:Opensim-dev at lists.berlios.de">Opensi=
m-dev at lists.berlios.de</a> <mailto:<a href=3D"mailto:Opensim-dev at lists.b=
erlios.de">Opensim-dev at lists.berlios.de</a>><br>
<div class=3D"im">>> <a href=3D"https://lists.berlios.de/mailman/list=
info/opensim-dev" target=3D"_blank">https://lists.berlios.de/mailman/listin=
fo/opensim-dev</a><br>
><br>
</div>> ----------------------------------------------------------------=
--------<br>
<div><div></div><div class=3D"h5">><br>
> _______________________________________________<br>
> Opensim-dev mailing list<br>
> <a href=3D"mailto:Opensim-dev at lists.berlios.de">Opensim-dev at lists.berl=
ios.de</a><br>
> <a href=3D"https://lists.berlios.de/mailman/listinfo/opensim-dev" targ=
et=3D"_blank">https://lists.berlios.de/mailman/listinfo/opensim-dev</a><br>
_______________________________________________<br>
Opensim-dev mailing list<br>
<a href=3D"mailto:Opensim-dev at lists.berlios.de">Opensim-dev at lists.berlios.d=
e</a><br>
<a href=3D"https://lists.berlios.de/mailman/listinfo/opensim-dev" target=3D=
"_blank">https://lists.berlios.de/mailman/listinfo/opensim-dev</a><br>
</div></div></blockquote></div><br></div>

--0016364ee01c638db7046ae1d243--