A bit confused

classic Classic list List threaded Threaded
20 messages Options
Reply | Threaded
Open this post in threaded view
|

A bit confused

Dr Ramesh Ramloll
Hello there, I have been enjoying my transition to opensim for a while and got something worthwhile I think going on. However, the past few experiences with the 0.7.4 is confusing me. I am experiencing loading delays, object contents not loading, clicks not detected ... both on local installs and on kitely. Are these issues new ... (introduced in 0.7.4) because the previous version seemed to be devoid of the problems I just mentioned. There was a thread that some experts I trust pointed me (regarding delays and so on 'slow handling of POST) but felt that the source of the problem is not yet known. Any clues to help me out?

--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: 208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Teravus Ovares
I'm not experiencing any of those at the moment.     I have seen some of the issues that you mention occur when there's certain kinds of loads on the simulator.    The touch detection could be a threading complication of some kind or it could be a communications issue with the simulator causing dropped events.  On the threading issue... I've seen this get particularly bad when the grid services are slow... but it has to do, generally with long running tasks that are supposed to be short... like web requests.   The 'contents' issue is probably a connection issue with the simulator...  that would get worse as more items are added to the contents. It could also be from malformed objects and inventories   I don't see that very often though.
 
Regards

Teravus
 


 
On Fri, Nov 2, 2012 at 10:17 AM, Dr Ramesh Ramloll <[hidden email]> wrote:
Hello there, I have been enjoying my transition to opensim for a while and got something worthwhile I think going on. However, the past few experiences with the 0.7.4 is confusing me. I am experiencing loading delays, object contents not loading, clicks not detected ... both on local installs and on kitely. Are these issues new ... (introduced in 0.7.4) because the previous version seemed to be devoid of the problems I just mentioned. There was a thread that some experts I trust pointed me (regarding delays and so on 'slow handling of POST) but felt that the source of the problem is not yet known. Any clues to help me out?

--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: <a href="tel:208-240-0040" target="_blank" value="+12082400040">208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Dr Ramesh Ramloll
Thanks Teravus for sharing ... hoping that the loading issues am facing are just something on my end.


On Sun, Nov 4, 2012 at 1:42 PM, Teravus Ovares <[hidden email]> wrote:
I'm not experiencing any of those at the moment.     I have seen some of the issues that you mention occur when there's certain kinds of loads on the simulator.    The touch detection could be a threading complication of some kind or it could be a communications issue with the simulator causing dropped events.  On the threading issue... I've seen this get particularly bad when the grid services are slow... but it has to do, generally with long running tasks that are supposed to be short... like web requests.   The 'contents' issue is probably a connection issue with the simulator...  that would get worse as more items are added to the contents. It could also be from malformed objects and inventories   I don't see that very often though.
 
Regards

Teravus
 


 
On Fri, Nov 2, 2012 at 10:17 AM, Dr Ramesh Ramloll <[hidden email]> wrote:
Hello there, I have been enjoying my transition to opensim for a while and got something worthwhile I think going on. However, the past few experiences with the 0.7.4 is confusing me. I am experiencing loading delays, object contents not loading, clicks not detected ... both on local installs and on kitely. Are these issues new ... (introduced in 0.7.4) because the previous version seemed to be devoid of the problems I just mentioned. There was a thread that some experts I trust pointed me (regarding delays and so on 'slow handling of POST) but felt that the source of the problem is not yet known. Any clues to help me out?

--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: <a href="tel:208-240-0040" value="+12082400040" target="_blank">208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users



--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: 208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Fleep Tuque
In reply to this post by Dr Ramesh Ramloll
We're running 0.7.4 on the University of Cincinnati grid and I'm also running it on FleepGrid and haven't experienced these kinds of delays, certainly nothing like lag in touch events or object contents loading, so maybe it's something in your environment?  

The only time I've seen that kind of behavior was when we were playing with physical enabled scripted prims and it was the physics slowing everything down.  Do you have physical objects enabled?  Does the behavior persist if you turn physics off?

- Chris/Fleep


Chris M. Collins (SL/OS: Fleep Tuque)
Center for Simulations & Virtual Environments Research (UCSIM)
UCIT Instructional & Research Computing
University of Cincinnati
406A Zimmer Hall
315 College Drive
PO BOX 210088
Cincinnati, OH 45221-0088
(513) 556-3018

http://ucsim.uc.edu

On Fri, Nov 2, 2012 at 11:17 AM, Dr Ramesh Ramloll <[hidden email]> wrote:
Hello there, I have been enjoying my transition to opensim for a while and got something worthwhile I think going on. However, the past few experiences with the 0.7.4 is confusing me. I am experiencing loading delays, object contents not loading, clicks not detected ... both on local installs and on kitely. Are these issues new ... (introduced in 0.7.4) because the previous version seemed to be devoid of the problems I just mentioned. There was a thread that some experts I trust pointed me (regarding delays and so on 'slow handling of POST) but felt that the source of the problem is not yet known. Any clues to help me out?

--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: <a href="tel:208-240-0040" value="+12082400040" target="_blank">208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Dr Ramesh Ramloll
Hi Chris,
I am avoiding physics the best I can (actually, we are not using physics at all right now). Usually I test things on a local opensim install before I upload finished stuff on kitely. So yes, at first I thought it may have have something to do with my own scripts. Then I tested with trivial stuff, like a prim with trivial scripts and textures on an empty sim, the behaviors persisted.. So now am thinking that may be something is wrong with my network set up ... (unlikely, because things were working fine before...and I made no changes). Will see, fingers crossed.
R


On Mon, Nov 5, 2012 at 9:35 AM, Fleep Tuque <[hidden email]> wrote:
We're running 0.7.4 on the University of Cincinnati grid and I'm also running it on FleepGrid and haven't experienced these kinds of delays, certainly nothing like lag in touch events or object contents loading, so maybe it's something in your environment?  

The only time I've seen that kind of behavior was when we were playing with physical enabled scripted prims and it was the physics slowing everything down.  Do you have physical objects enabled?  Does the behavior persist if you turn physics off?

- Chris/Fleep


Chris M. Collins (SL/OS: Fleep Tuque)
Center for Simulations & Virtual Environments Research (UCSIM)
UCIT Instructional & Research Computing
University of Cincinnati
406A Zimmer Hall
315 College Drive
PO BOX 210088
Cincinnati, OH 45221-0088
<a href="tel:%28513%29%20556-3018" value="+15135563018" target="_blank">(513) 556-3018

http://ucsim.uc.edu

On Fri, Nov 2, 2012 at 11:17 AM, Dr Ramesh Ramloll <[hidden email]> wrote:
Hello there, I have been enjoying my transition to opensim for a while and got something worthwhile I think going on. However, the past few experiences with the 0.7.4 is confusing me. I am experiencing loading delays, object contents not loading, clicks not detected ... both on local installs and on kitely. Are these issues new ... (introduced in 0.7.4) because the previous version seemed to be devoid of the problems I just mentioned. There was a thread that some experts I trust pointed me (regarding delays and so on 'slow handling of POST) but felt that the source of the problem is not yet known. Any clues to help me out?

--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: <a href="tel:208-240-0040" value="+12082400040" target="_blank">208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users



--
'Consider how the lilies grow. They do not labor or spin.'
Rameshsharma Ramloll PhD, CEO CTO DeepSemaphore LLC, Affiliate Research Associate Professor, Idaho State University, Pocatello, ID 83209 Tel: 208-240-0040


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

justincc
Hopefully, this week I'm going to open a mantis report about this issue (0.7.4 suffering comms issues compared to
0.7.3).  Enough data has been reported now where I suspect it's an issue that isn't network or configuration related
(though that's still a possibility).

I will probably end up starting a wiki page about all the diagnostic steps for this kind of thing, but one immediate
thing to try is to disable HTTP inventory CAPs, as these were on by default for 0.7.4 but not for 0.7.3.  You can do
this by setting

[ClientStack.LindenCaps]
     Cap_FetchInventoryDescendents2 = ""
     Cap_FetchInventory2 = ""

in OpenSim.ini to override the localhost settings in OpenSimDefaults.ini.  This will force viewers to fallback on the
older UDP inventory mechanisms.  But beware - this fallback was broken in very recent (3.3.x) LL viewers and I'm not
sure what the state of play is for TPVs.

I'm particularly interested in this setting since I'm beginning to suspect the issue is in this area due to the sheer
number of inventory requests made by some viewers (as shown by Ai Austin's previous posts).  Indeed, I think that I
found an issue with OpenSimulator's inventory folder version updating (or lack thereof) today.  It's possible (though
this is speculative at this point), that this issue is causing viewers to re-request a lot of inventory information
where they could actually reuse their cached information.  In theory, that should affect UDP inventory requests as well,
though these do not tie up TCP connections in the same way as inventory CAPs.

However, it's also worth bearing in mind that OSGrid has been undergoing an asset migration recently that very
significantly degraded AssetService responsiveness at times.  This could be contributing to Ai's problems and yours if
you are on osgrid (though I vaguely remember he said that his own grid was suffereing as well).

On 05/11/12 14:52, Dr Ramesh Ramloll wrote:

> Hi Chris,
> I am avoiding physics the best I can (actually, we are not using physics at all right now). Usually I test things on a
> local opensim install before I upload finished stuff on kitely. So yes, at first I thought it may have have something to
> do with my own scripts. Then I tested with trivial stuff, like a prim with trivial scripts and textures on an empty sim,
> the behaviors persisted.. So now am thinking that may be something is wrong with my network set up ... (unlikely,
> because things were working fine before...and I made no changes). Will see, fingers crossed.
> R
>
>
> On Mon, Nov 5, 2012 at 9:35 AM, Fleep Tuque <[hidden email] <mailto:[hidden email]>> wrote:
>
>     We're running 0.7.4 on the University of Cincinnati grid and I'm also running it on FleepGrid and haven't
>     experienced these kinds of delays, certainly nothing like lag in touch events or object contents loading, so maybe
>     it's something in your environment?
>
>     The only time I've seen that kind of behavior was when we were playing with physical enabled scripted prims and it
>     was the physics slowing everything down.  Do you have physical objects enabled?  Does the behavior persist if you
>     turn physics off?
>
>     - Chris/Fleep
>
>
>     Chris M. Collins (SL/OS: Fleep Tuque)
>     Center for Simulations & Virtual Environments Research (UCSIM)
>     UCIT Instructional & Research Computing
>     University of Cincinnati
>     406A Zimmer Hall
>     315 College Drive
>     PO BOX 210088
>     Cincinnati, OH 45221-0088
>     [hidden email] <mailto:[hidden email]>
>     (513) 556-3018 <tel:%28513%29%20556-3018>
>
>     http://ucsim.uc.edu
>
>     On Fri, Nov 2, 2012 at 11:17 AM, Dr Ramesh Ramloll <[hidden email] <mailto:[hidden email]>> wrote:
>
>         Hello there, I have been enjoying my transition to opensim for a while and got something worthwhile I think
>         going on. However, the past few experiences with the 0.7.4 is confusing me. I am experiencing loading delays,
>         object contents not loading, clicks not detected ... both on local installs and on kitely. Are these issues new
>         ... (introduced in 0.7.4) because the previous version seemed to be devoid of the problems I just mentioned.
>         There was a thread that some experts I trust pointed me (regarding delays and so on 'slow handling of POST) but
>         felt that the source of the problem is not yet known. Any clues to help me out?
>
>         --
>         'Consider how the lilies grow. They do not labor or spin.'
>         *Rameshsharma Ramloll* PhD, CEO CTO DeepSemaphore LLC, Affiliate /Research Associate Professor/, Idaho State
>         University, Pocatello, ID 83209 Tel: 208-240-0040 <tel:208-240-0040>
>         Blog <http://deepsemaphore.posterous.com/>, LinkedIn <http://www.linkedin.com/in/rameshramloll>, DeepSemaphore
>         LLC <http://www.deepsemaphore.com>, Google+ profile <https://plus.google.com/103652369558830540272/about>
>
>
>         _______________________________________________
>         Opensim-users mailing list
>         [hidden email] <mailto:[hidden email]>
>         https://lists.berlios.de/mailman/listinfo/opensim-users
>
>
>
>     _______________________________________________
>     Opensim-users mailing list
>     [hidden email] <mailto:[hidden email]>
>     https://lists.berlios.de/mailman/listinfo/opensim-users
>
>
>
>
> --
> 'Consider how the lilies grow. They do not labor or spin.'
> *Rameshsharma Ramloll* PhD, CEO CTO DeepSemaphore LLC, Affiliate /Research Associate Professor/, Idaho State University,
> Pocatello, ID 83209 Tel: 208-240-0040
> Blog <http://deepsemaphore.posterous.com/>, LinkedIn <http://www.linkedin.com/in/rameshramloll>, DeepSemaphore LLC
> <http://www.deepsemaphore.com>, Google+ profile <https://plus.google.com/103652369558830540272/about>
>
>
>
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/opensim-users
>


--
Justin Clark-Casey (justincc)
OSVW Consulting
http://justincc.org
http://twitter.com/justincc
_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

aiaustin
In reply to this post by Dr Ramesh Ramloll
At 11:00 07/11/2012, Justin Clark-Casey <[hidden email]> wrote:
>Hopefully, this week I'm going to open a mantis report about this
>issue (0.7.4 suffering comms issues compared to
>0.7.3). Enough data has been reported now where I suspect it's an
>issue that isn't network or configuration related
>(though that's still a possibility).

I am not surprised by the title of this post... as my own situation
in reporting the issues all coincided with me switching to using
Firestorm (and Dolphin a bit) whereas before that I had recently only
used almost exclusively LL official 3.x  viewers.  I also did a
revamp of our OSGrid add on 0.7.5 dev regions to switch to a new
dedicated host with MySQL for its DB rather than SQLite too.  So its
difficult to know what the cause is or causes are... but something is
definitely not quite right and the issue(s) can really seriously jam
up regions even when a sole user has logged off for some time, or a
single user is trying to get back in after a jam up.

It sounds like those of us using the most recent viewers based on LL
3.3.x just before it stopped supporting OpenSim may not be getting
helpful results to report if the HTTP inventory CAPS fall back
mechanism is also broken in some TPVs based on LL 3.3.x code (like
Firestorm and Dolphin?)

I will try to watch for recurrent symptoms and report what I can as
the Mantis entry is filed and others home in on the issue.


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Oren Hurvitz
In reply to this post by justincc
I've been looking at this issue as well, and it's pretty certain that it's due to the number of HTTP inventory requests sent from the viewer. I added logging that shows that when the number of incoming HTTP requests to OpenSim exceeds about 10/second it starts handling requests very slowly (i.e., handling 50 requests at once is slower than handling 50 requests that come in one at a time). I was actually able to improve performance a bit by adding delays to make OpenSim accept incoming HTTP connections more slowly! But that's a dirty hack, and I'm looking for a better solution.

I have found that one reliable source of problems happens when a user logs-in: the LLLoginResponse packet includes a list of all of the user's active gestures, and then the viewer sends *separate* FetchInventory requests for each of those gestures. For example, if the user has 50 active gestures then the viewer sends 50 HTTP requests at once, and that overwhelms OpenSim for a while. During that time other requests from the client are also handled slowly.

I've put in a request to Firestorm (as the leading OpenSim viewer) to change their behavior so that they'll request all of the gestures at once: see http://jira.phoenixviewer.com/browse/FIRE-8146 . But they've put this issue on the "Wishlist", so unless someone steps in (Armin Weatherwax?) this might never be fixed on the viewer side.
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Opensimfan
In reply to this post by justincc


question about this
[ClientStack.LindenCaps]
     Cap_FetchInventoryDescendents2 = ""
     Cap_FetchInventory2 = "" 
would it be possible to point this to grid sever? something like this...

EXAMPLE..!!
[ClientStack.LindenCaps]
     Cap_FetchInventoryDescendents2 = "http://inventory.osgrid.org"
     Cap_FetchInventory2 = "http://inventory.osgrid.org"

André
_________________________________________
OpenSimFan

My Opensim/Second Life Blog
http://verwijs.wordpress.com

(Dutch, basic hardware/software help windows, Mac, Linux)
http://verwijs-pc.nl

My Twitter Page:
http://twitter.com/OpenSimFan

My Facebook page (be my friend, please )
http://www.facebook.com/andre.verwijs

My Google+ page (follow me please )
André Verwijs - Google+
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Diva Canto
Yes, of course. That's why those config vars are there. That won't solve the problem of slow posts, though; will it continue to overwhelm the inventory server, this time directly from the viewers. But at least the sim will be out of the loop.

On 11/7/2012 11:32 AM, OpenSimFan wrote:


question about this
[ClientStack.LindenCaps]
     Cap_FetchInventoryDescendents2 = ""
     Cap_FetchInventory2 = "" 
would it be possible to point this to grid sever? something like this...

EXAMPLE..!!
[ClientStack.LindenCaps]
     Cap_FetchInventoryDescendents2 = "http://inventory.osgrid.org"
     Cap_FetchInventory2 = "http://inventory.osgrid.org"

André
_________________________________________
OpenSimFan

My Opensim/Second Life Blog
http://verwijs.wordpress.com

(Dutch, basic hardware/software help windows, Mac, Linux)
http://verwijs-pc.nl

My Twitter Page:
http://twitter.com/OpenSimFan

My Facebook page (be my friend, please )
http://www.facebook.com/andre.verwijs



View this message in context: Re: A bit confused
Sent from the opensim-users mailing list archive at Nabble.com.


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

M.E. Verhagen
In reply to this post by Oren Hurvitz
The sending of the fetch request by firestorm sounds a bit strange, since opensim already knows these gestures have to be send to the viewer. Can't opensim just send those gestures and ignore those fetch requests ?

Op 7 nov. 2012 om 17:36 heeft Oren Hurvitz <[hidden email]> het volgende geschreven:

> I've been looking at this issue as well, and it's pretty certain that it's
> due to the number of HTTP inventory requests sent from the viewer. I added
> logging that shows that when the number of incoming HTTP requests to OpenSim
> exceeds about 10/second it starts handling requests very slowly (i.e.,
> handling 50 requests at once is slower than handling 50 requests that come
> in one at a time). I was actually able to improve performance a bit by
> adding delays to make OpenSim accept incoming HTTP connections more slowly!
> But that's a dirty hack, and I'm looking for a better solution.
>
> I have found that one reliable source of problems happens when a user
> logs-in: the LLLoginResponse packet includes a list of all of the user's
> active gestures, and then the viewer sends *separate* FetchInventory
> requests for each of those gestures. For example, if the user has 50 active
> gestures then the viewer sends 50 HTTP requests at once, and that overwhelms
> OpenSim for a while. During that time other requests from the client are
> also handled slowly.
>
> I've put in a request to Firestorm (as the leading OpenSim viewer) to change
> their behavior so that they'll request all of the gestures at once: see
> http://jira.phoenixviewer.com/browse/FIRE-8146 . But they've put this issue
> on the "Wishlist", so unless someone steps in (Armin Weatherwax?) this might
> never be fixed on the viewer side.
>
>
>
> --
> View this message in context: http://opensim-users.2152040.n2.nabble.com/A-bit-confused-tp7578881p7578896.html
> Sent from the opensim-users mailing list archive at Nabble.com.
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/opensim-users
_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

justincc
It cannot.  Inventory requests can only be initiated viewer-side.

On 08/11/12 00:00, Marcel Verhagen wrote:

> The sending of the fetch request by firestorm sounds a bit strange, since opensim already knows these gestures have to be send to the viewer. Can't opensim just send those gestures and ignore those fetch requests ?
>
> Op 7 nov. 2012 om 17:36 heeft Oren Hurvitz <[hidden email]> het volgende geschreven:
>
>> I've been looking at this issue as well, and it's pretty certain that it's
>> due to the number of HTTP inventory requests sent from the viewer. I added
>> logging that shows that when the number of incoming HTTP requests to OpenSim
>> exceeds about 10/second it starts handling requests very slowly (i.e.,
>> handling 50 requests at once is slower than handling 50 requests that come
>> in one at a time). I was actually able to improve performance a bit by
>> adding delays to make OpenSim accept incoming HTTP connections more slowly!
>> But that's a dirty hack, and I'm looking for a better solution.
>>
>> I have found that one reliable source of problems happens when a user
>> logs-in: the LLLoginResponse packet includes a list of all of the user's
>> active gestures, and then the viewer sends *separate* FetchInventory
>> requests for each of those gestures. For example, if the user has 50 active
>> gestures then the viewer sends 50 HTTP requests at once, and that overwhelms
>> OpenSim for a while. During that time other requests from the client are
>> also handled slowly.
>>
>> I've put in a request to Firestorm (as the leading OpenSim viewer) to change
>> their behavior so that they'll request all of the gestures at once: see
>> http://jira.phoenixviewer.com/browse/FIRE-8146 . But they've put this issue
>> on the "Wishlist", so unless someone steps in (Armin Weatherwax?) this might
>> never be fixed on the viewer side.
>>
>>
>>
>> --
>> View this message in context: http://opensim-users.2152040.n2.nabble.com/A-bit-confused-tp7578881p7578896.html
>> Sent from the opensim-users mailing list archive at Nabble.com.
>> _______________________________________________
>> Opensim-users mailing list
>> [hidden email]
>> https://lists.berlios.de/mailman/listinfo/opensim-users
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/opensim-users
>


--
Justin Clark-Casey (justincc)
OSVW Consulting
http://justincc.org
http://twitter.com/justincc
_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

justincc
In reply to this post by aiaustin
On 07/11/12 11:19, Ai Austin wrote:

> At 11:00 07/11/2012, Justin Clark-Casey <[hidden email]> wrote:
>> Hopefully, this week I'm going to open a mantis report about this issue (0.7.4 suffering comms issues compared to
>> 0.7.3). Enough data has been reported now where I suspect it's an issue that isn't network or configuration related
>> (though that's still a possibility).
>
> I am not surprised by the title of this post... as my own situation in reporting the issues all coincided with me
> switching to using Firestorm (and Dolphin a bit) whereas before that I had recently only used almost exclusively LL
> official 3.x  viewers.  I also did a revamp of our OSGrid add on 0.7.5 dev regions to switch to a new dedicated host
> with MySQL for its DB rather than SQLite too.  So its difficult to know what the cause is or causes are... but something
> is definitely not quite right and the issue(s) can really seriously jam up regions even when a sole user has logged off
> for some time, or a single user is trying to get back in after a jam up.
>
> It sounds like those of us using the most recent viewers based on LL 3.3.x just before it stopped supporting OpenSim may
> not be getting helpful results to report if the HTTP inventory CAPS fall back mechanism is also broken in some TPVs
> based on LL 3.3.x code (like Firestorm and Dolphin?)
>
> I will try to watch for recurrent symptoms and report what I can as the Mantis entry is filed and others home in on the
> issue.

Okay, I have now created Mantis 6412 [1] for investigation of this issue.

I have just today made some fixes so that updates to folder inventory versions stored by OpenSimulator now match those
assumed by the viewers and stored in its cache.  Previously, version folder number was only correctly incremented on
item add, whereas the viewer was also expecting updates on item move/remove and folder create/remove/move.  If the
folder version stored by the viewer was greater than that reported by the simulator, the viewer would dump its cached
information.

Now that the numbers match, the viewer no longer dumps this information.  I'm hoping that this will lead to much more
use of the inventory cache viewer-side and reduce the number of inventory requests.  It also appears to resolve an issue
where updates of the "Current Outfit" folder via external mechanisms (or perhaps another viewer) would not show up in
the viewer until 2 relogs or a cache clear.

So as I know you follow the master code, Ai, could you test this and see if it has an impact?  This will need a ROBUST
update, and so will not be live on osgrid until the services are updated there, but if I remember you've also seen this
on your own grid.  To have an effect, as above, it may require 2 viewer relogs and/or a cache clear.

[1] http://opensimulator.org/mantis/view.php?id=6412

--
Justin Clark-Casey (justincc)
OSVW Consulting
http://justincc.org
http://twitter.com/justincc
_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

justincc
In reply to this post by Oren Hurvitz
Yeah, that's not good.  I suspect the C# HttpServer we're using is not very performant but I'm not aware of any testing
done on that.  And as you probably know, there are few alternatives short of directing the CAP directly to a webserver
for processing (which would be a lot of work though and interesting long-term solution for larger installations).

I'm hoping that correcting inventory version numbers will improve matters though that's largely a hypothesis at this
stage.  Also, if a cache-cleared viewer is determined to launch large number of requests at once that would still be a
problem.

I wonder if this kind of issue is behind LL's work on improving CAP comms.

On 07/11/12 16:36, Oren Hurvitz wrote:

> I've been looking at this issue as well, and it's pretty certain that it's
> due to the number of HTTP inventory requests sent from the viewer. I added
> logging that shows that when the number of incoming HTTP requests to OpenSim
> exceeds about 10/second it starts handling requests very slowly (i.e.,
> handling 50 requests at once is slower than handling 50 requests that come
> in one at a time). I was actually able to improve performance a bit by
> adding delays to make OpenSim accept incoming HTTP connections more slowly!
> But that's a dirty hack, and I'm looking for a better solution.
>
> I have found that one reliable source of problems happens when a user
> logs-in: the LLLoginResponse packet includes a list of all of the user's
> active gestures, and then the viewer sends *separate* FetchInventory
> requests for each of those gestures. For example, if the user has 50 active
> gestures then the viewer sends 50 HTTP requests at once, and that overwhelms
> OpenSim for a while. During that time other requests from the client are
> also handled slowly.
>
> I've put in a request to Firestorm (as the leading OpenSim viewer) to change
> their behavior so that they'll request all of the gestures at once: see
> http://jira.phoenixviewer.com/browse/FIRE-8146 . But they've put this issue
> on the "Wishlist", so unless someone steps in (Armin Weatherwax?) this might
> never be fixed on the viewer side.
>
>
>
> --
> View this message in context: http://opensim-users.2152040.n2.nabble.com/A-bit-confused-tp7578881p7578896.html
> Sent from the opensim-users mailing list archive at Nabble.com.
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/opensim-users
>


--
Justin Clark-Casey (justincc)
OSVW Consulting
http://justincc.org
http://twitter.com/justincc
_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

aiaustin
In reply to this post by Dr Ramesh Ramloll
At 11:00 09/11/2012, JustinCC wrote:
>So as I know you follow the master code, Ai, could you test this and
>see if it has an impact?  This will need a ROBUST
>update, and so will not be live on osgrid until the services are
>updated there, but if I remember you've also seen this
>on your own grid.  To have an effect, as above, it may require 2
>viewer relogs and/or a cache clear.

http://opensimulator.org/mantis/view.php?id=6412

Justin, I have updated the Openvue grid to r/20937 9-Nov-2012 and
checked we are all set for Diva's HG2.0 config changes at the same
time.   I also cleared all my own viewer caches just to be sure.

I am not actually sure I have seen the serious issues on Openvue grid
- or not so obviously anyway... it seems to be our own add on regions
to OSGrid that are affected, which is what I have reported.  I will
update those now too,but as you say we need to wait fro a core Robust
services update on OSGrid to have this in a clean state for testing
and reporting if there are issues remaining after your
changes.  Maybe the OSGrid folks could let us know via this list when
that is updated so we know when to recheck things?


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Michael Emory Cerquoni
The OSgrid Robust server has the Diva HG 2.0 updates but because of the configuration of OSgrid it will not provide any benefit for this grid, because of the open nature of the grid we can not really benefit from these changes.  Is there any specific Git HASH that you were looking for specifically, I can verify we are fully updated.

On Fri, Nov 9, 2012 at 6:49 AM, Ai Austin <[hidden email]> wrote:
At 11:00 09/11/2012, JustinCC wrote:
So as I know you follow the master code, Ai, could you test this and see if it has an impact?  This will need a ROBUST
update, and so will not be live on osgrid until the services are updated there, but if I remember you've also seen this
on your own grid.  To have an effect, as above, it may require 2 viewer relogs and/or a cache clear.

http://opensimulator.org/mantis/view.php?id=6412

Justin, I have updated the Openvue grid to r/20937 9-Nov-2012 and checked we are all set for Diva's HG2.0 config changes at the same time.   I also cleared all my own viewer caches just to be sure.

I am not actually sure I have seen the serious issues on Openvue grid - or not so obviously anyway... it seems to be our own add on regions to OSGrid that are affected, which is what I have reported.  I will update those now too,but as you say we need to wait fro a core Robust services update on OSGrid to have this in a clean state for testing and reporting if there are issues remaining after your changes.  Maybe the OSGrid folks could let us know via this list when that is updated so we know when to recheck things?



_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users



--
Michael Emory Cerquoni

_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

Diva Canto
HG 2.0 and HG 1.5 work seamlessly with each other. The configurations are entirely under the control of the grid operators, who can choose what works best for their grid.
Grids may continue to use the HG 1.5 inventory service if they want. They may even use the regular (HG 1.0) inventory service in some cases.

In the case of OSGrid, they are using the regular (HG 1.0) inventory service for the reasons that Nebadon points out.** I don't recommend it in the strongest possible way.** Only OSGrid, as testing grid, justifies keeping inventory like that. The other exception for which it may be ok is when the interoperating grids form a self-contained network where trust is ensured by external means and no one ever goes outside of that network.

On 11/9/2012 6:46 AM, Nebadon Izumi wrote:
The OSgrid Robust server has the Diva HG 2.0 updates but because of the configuration of OSgrid it will not provide any benefit for this grid, because of the open nature of the grid we can not really benefit from these changes.  Is there any specific Git HASH that you were looking for specifically, I can verify we are fully updated.

On Fri, Nov 9, 2012 at 6:49 AM, Ai Austin <[hidden email]> wrote:
At 11:00 09/11/2012, JustinCC wrote:
So as I know you follow the master code, Ai, could you test this and see if it has an impact?  This will need a ROBUST
update, and so will not be live on osgrid until the services are updated there, but if I remember you've also seen this
on your own grid.  To have an effect, as above, it may require 2 viewer relogs and/or a cache clear.

http://opensimulator.org/mantis/view.php?id=6412

Justin, I have updated the Openvue grid to r/20937 9-Nov-2012 and checked we are all set for Diva's HG2.0 config changes at the same time.   I also cleared all my own viewer caches just to be sure.

I am not actually sure I have seen the serious issues on Openvue grid - or not so obviously anyway... it seems to be our own add on regions to OSGrid that are affected, which is what I have reported.  I will update those now too,but as you say we need to wait fro a core Robust services update on OSGrid to have this in a clean state for testing and reporting if there are issues remaining after your changes.  Maybe the OSGrid folks could let us know via this list when that is updated so we know when to recheck things?



_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users



--
Michael Emory Cerquoni


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

justincc
In reply to this post by Michael Emory Cerquoni
Require commit is 61b816ebba57307410d877c4190e81b6469718cb from Fri Nov 9 01:18:11 2012

If you only see it on osgrid then this is an important point.  As we know, asset migration has/is going on which make
effect service response times and there are many other possible factors.  What I really need is data from smaller grids,
preferably as simple as possible (e.g. all simulators in same place with services).

On 09/11/12 14:46, Nebadon Izumi wrote:

> The OSgrid Robust server has the Diva HG 2.0 updates but because of the configuration of OSgrid it will not provide any
> benefit for this grid, because of the open nature of the grid we can not really benefit from these changes.  Is there
> any specific Git HASH that you were looking for specifically, I can verify we are fully updated.
>
> On Fri, Nov 9, 2012 at 6:49 AM, Ai Austin <[hidden email] <mailto:[hidden email]>> wrote:
>
>     At 11:00 09/11/2012, JustinCC wrote:
>
>         So as I know you follow the master code, Ai, could you test this and see if it has an impact?  This will need a
>         ROBUST
>         update, and so will not be live on osgrid until the services are updated there, but if I remember you've also
>         seen this
>         on your own grid.  To have an effect, as above, it may require 2 viewer relogs and/or a cache clear.
>
>
>     http://opensimulator.org/__mantis/view.php?id=6412 <http://opensimulator.org/mantis/view.php?id=6412>
>
>     Justin, I have updated the Openvue grid to r/20937 9-Nov-2012 and checked we are all set for Diva's HG2.0 config
>     changes at the same time.   I also cleared all my own viewer caches just to be sure.
>
>     I am not actually sure I have seen the serious issues on Openvue grid - or not so obviously anyway... it seems to be
>     our own add on regions to OSGrid that are affected, which is what I have reported.  I will update those now too,but
>     as you say we need to wait fro a core Robust services update on OSGrid to have this in a clean state for testing and
>     reporting if there are issues remaining after your changes.  Maybe the OSGrid folks could let us know via this list
>     when that is updated so we know when to recheck things?
>
>
>
>     _________________________________________________
>     Opensim-users mailing list
>     [hidden email] <mailto:[hidden email]>
>     https://lists.berlios.de/__mailman/listinfo/opensim-users <https://lists.berlios.de/mailman/listinfo/opensim-users>
>
>
>
>
> --
> Michael Emory Cerquoni
>
>
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/opensim-users
>


--
Justin Clark-Casey (justincc)
OSVW Consulting
http://justincc.org
http://twitter.com/justincc
_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

aiaustin
In reply to this post by Dr Ramesh Ramloll

>From: Nebadon Izumi <[hidden email]>
>Is there any specific Git HASH that you were looking for
>specifically, I can verify we are fully updated.

Thanks Nebadon.. as Justin mentioned we want to know when the core
Robust services on OSGrid are at r/20937
(61b816ebba57307410d877c4190e81b6469718cb from Fri Nov 9 01:18:11
2012) as then we can give updated symptom reports to Justin.

Required commit is

If you can at the same time do a proper OSGrid add-on region upadte
to or past that release then that will simplify deployment and
testing... but I can update my OSGrid Vue addon regions to a 0.7.5
GIT master version and merge with the required OSGrid config files
for testing if necessary, as we have done that a few times for
specific testing for Justin before.

For the avoidance of doubt, I do not think this problem relates at
all to HG changes made recently by Diva.... as the problems occurred
before the change to HG2.0 anyway.

_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: A bit confused

aiaustin
In reply to this post by Dr Ramesh Ramloll
At 09:55 10/11/2012, Ai Austin wrote:
>If you can at the same time do a proper OSGrid add-on region update
>to or past that release then that will simplify deployment and testing...


Ah, I have just seen that you are ahead of me on that and there is a
release there now.. OpenSimulator v0.7.5.4f98259 ... I will install
and get in sync shortly...


_______________________________________________
Opensim-users mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/opensim-users