First thing i asking myself, how could things be rezzed on HG region.
Its not smart to allow any scripts or objects to run or rezz.
Its anyway asking for problems to keep that enabled.
If you run linux, there some tools like monit that show you in
webinterface at 1 clane if there's something wrong with cpu or memory of
a region.
You only need to have regions split 1:1 what i do for years. also if 1
region crash you dont tear others down.
happy you found the problem
On 2012-02-21 23:48, Shaun Erickson wrote:
Salad Boy strikes again! He did the same thing to me a few months ago.
-ste
On Feb 21, 2012, at 5:28 PM, Fleep Tuque <[email protected]
<mailto:[email protected]>> wrote:
Hi all,
If any of you were following the long email thread on opensim-users
about intermittent crashing and out of memory exceptions, you may
recall I just figured out how to spread regions across multiple
opensim instances. (If you missed the thread, see
http://opensim-users.2152040.n2.nabble.com/Intermittent-crashing-System-OutofMemoryException-td7281298.html.)
So today, the HGAC folks came to visit FleepGrid and things went
relatively well, and after it was over I thought I'd follow Rick's
suggestion and load each sim one at a time to see how much memory
they used and general performance for each region. Considering
FleepGrid is running on an old P4 with 4GB RAM, I've been relatively
pleased with its performance with around 11 regions, some of them
with lots of content and scripts. The only major concern has been
the CPU consistently pegging around 75-80% usage after all the
scripts are loaded (it always hits 100% during start up).
Imagine my surprise when I loaded each sim one at a time and was
seeing virtually no CPU load after the scripts finished loading,
region after region, even when I started adding them back
cumulatively, CPU load was hovering around 2%! What the heck was
going on, I wondered, the only regions I had left to load were my
three hypergate regions which are basically small islands with a few
signs and very few scripts or objects.
Lo and behold, I added HGate 1000 and suddenly the CPU load shot back
up to 78% after scripts finished loading. I couldn't imagine what
the heck was going on with that nearly empty region but upon
investigation, I discovered 600 spheres set to physical under the
waterline, each taking up precious CPU resources.
The long and short of it is, I hadn't checked those regions in ages
because they're just simple hypergate jump points and I didn't think
anything in particular was going on there, but those darn physical
balls have been chewing up my CPU (and making the $#%^@$ loud fan run
all day and night) for goodness knows how long, PLUS whatever
performance hit the whole grid suffered as a result of my
inattentiveness. Hopefully this will mean a lot less laggy
experience on FleepGrid and a lot less stress on my poor old PC
(which I'm completely surprised was even able to handle 600 physical
objects at once!).
So just a friendly reminder to myself and you to check your sims
periodically for griefer crap. And curses to user "Jack Marioline"
wherever you are. :P
- Chris/Fleep
_______________________________________________
Opensim-users mailing list
[email protected] <mailto:[email protected]>
https://lists.berlios.de/mailman/listinfo/opensim-users
_______________________________________________
Opensim-users mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/opensim-users
_______________________________________________
Opensim-users mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/opensim-users
