consider commercial support from consulting@freeswitch.org On Fri, Feb 12, 2010 at 10:33 AM, Joel Sisko <joel.sisko@iconverged.com>wrote:

Brian/Gavin do you know of any resource/link that can give an indication on what we could be expected of conference capacity? We are trying to determine a few different platforms we can possible use for conferencing and doing a little preliminary homework.

Kindest regards,

Joel

----- Original Message ----- From: "Joel Sisko" <joel.sisko@iconverged.com> To: freeswitch-users@lists.freeswitch.org Sent: Thursday, February 11, 2010 4:05:58 PM GMT -08:00 US/Canada Pacific Subject: Re: [Freeswitch-users] Mod_Conference capacity....

Brian/Gavin thanks for the input. But I agree with Brian, if were that easy I would have done it prior to the post.

Just looking to find out what some of the communities success has been to see if this is a path we should go down for a conference solution platform.

Joel ----- Original Message ----- From: "Brian West" <brian@freeswitch.org> To: freeswitch-users@lists.freeswitch.org Sent: Thursday, February 11, 2010 3:47:46 PM GMT -08:00 US/Canada Pacific Subject: Re: [Freeswitch-users] Mod_Conference capacity....

Not an optimal test scenario unless you know wtf you're doing!

/b

On Feb 11, 2010, at 5:22 PM, Gavin Henry wrote:

...

Why don't use script a test or use sipp and then dial in yourself to listen?

Cheers.

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:+19193869900

Producing benchmark numbers for an opensource project is a cardinal sin. I can safely tell you that it is "many" and very competitive with anything else you will encounter as long as you use a modern multi-core 64bit machine. As I said there is commercial support available which is customary for anyone needing assistance setting up a company. On Fri, Feb 12, 2010 at 11:30 AM, Joel Sisko <joel.sisko@iconverged.com>wrote:

So if I understand correctly that there are no numbers to be considered from anyone at this point in regards to capacity as a general guideline?

Kindest regards,

Joel ----- Original Message ----- From: "Anthony Minessale" <anthony.minessale@gmail.com> To: freeswitch-users@lists.freeswitch.org Sent: Friday, February 12, 2010 8:56:31 AM GMT -08:00 US/Canada Pacific Subject: Re: [Freeswitch-users] Mod_Conference capacity....

consider commercial support from consulting@freeswitch.org

On Fri, Feb 12, 2010 at 10:33 AM, Joel Sisko <joel.sisko@iconverged.com>wrote:

...

Brian/Gavin do you know of any resource/link that can give an indication on what we could be expected of conference capacity? We are trying to determine a few different platforms we can possible use for conferencing and doing a little preliminary homework.

Kindest regards,

Joel

----- Original Message ----- From: "Joel Sisko" <joel.sisko@iconverged.com> To: freeswitch-users@lists.freeswitch.org Sent: Thursday, February 11, 2010 4:05:58 PM GMT -08:00 US/Canada Pacific Subject: Re: [Freeswitch-users] Mod_Conference capacity....

Brian/Gavin thanks for the input. But I agree with Brian, if were that easy I would have done it prior to the post.

Just looking to find out what some of the communities success has been to see if this is a path we should go down for a conference solution platform.

Joel ----- Original Message ----- From: "Brian West" <brian@freeswitch.org> To: freeswitch-users@lists.freeswitch.org Sent: Thursday, February 11, 2010 3:47:46 PM GMT -08:00 US/Canada Pacific Subject: Re: [Freeswitch-users] Mod_Conference capacity....

Not an optimal test scenario unless you know wtf you're doing!

/b

On Feb 11, 2010, at 5:22 PM, Gavin Henry wrote:

...

Why don't use script a test or use sipp and then dial in yourself to listen?

Cheers.

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

-- Anthony Minessale II

FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire

AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch

FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:+19193869900

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE: http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:+19193869900

On Fri, Feb 12, 2010 at 11:36 PM, Gavin Henry <gavin.henry@gmail.com> wrote:

Hi,

I think this is unfair to automatically point a user to commercial support. Its like holding your users hostage until they pay for info.

well, the point is that every application is different and nobody can say what the performance for this or that HW exactly is. This is not like holding a hostage it is more like "we need to play on your existing platform run some tests so we can come out with some real benchmark"... this is what i can read from Anthony's e-mail. I'm willing to write a page on the wiki with info on the recommended

tools to use etc. with links to the commercial support resource list if they can't be bothered to do the setup themselves or want the experts paid for guidance.

yes, SIPP together with nmon and wireshark and Adobe Audition are the right tools (at least thats what i'm using...) for such benchmarking. - SIPP to generate traffic load, - nmon to get some real stats on the machine - wireshark to quickly check the jitter on your test call and extract the voice stream (sniffing has to be done on a mirrored switch port) - adobe autition to perform voice quality check on the extracted voice stream

It just reads like a bit of an insult. Have a question? Well you must be stupid and need to pay for help immediately.

dont think so, "you must learn how to use all these tools"... if you don't want to do it, there is another option but it is not free :(

He may have even written the page for the docs team if he was encouraged first with a few pointers on how to do the tests and written them up with results for others to find with a big YMMV warning.

well ... i think the time is always the issue + nobody likes documenting things :(

Just my thoughts from experiences as the Doc team lead for the OpenLDAP project and dealing with lots of users asking questions too.

But some users never read docs and there a very good amount of high quality pages in the wiki. Others, if helped, come back and contribute, but you may have lost this one.

I cannot say anything here except you are right and i agree what you are saying about documenting things ...

Gav.

On 13 February 2010 15:22, Anthony Minessale <anthony.minessale@gmail.com> wrote:

Look through the archives of this list and count the answers to questions I provide daily for the last several years.  It will be a clear pattern.

People persistantly asking about load testing and benchmark numbers will always get the same response: we only support it commerically.

We learned from experience the dangers of entertaining such questioning and its our policy to not do so over our free public forum.

OK, thanks for the info. If this is now the case these three links should be updated to reflect the FreeSWITCH projects stand point: http://wiki.freeswitch.org/wiki/Performance_testing_and_configurations http://wiki.freeswitch.org/wiki/Long_term_testing_chevymanjosh http://wiki.freeswitch.org/wiki/FreeSwitch_FAQ#Q:_How_many_concurrent_calls_... or a new FAQ entry created etc. Thanks, Gavin. -- http://www.suretecsystems.com/services/openldap/ http://www.suretectelecom.com

On 13 February 2010 19:54, Brian West <brian@freeswitch.org> wrote:

This is the FAQ entry.  It applies to EVERYTHING related to load testing. /b On Feb 13, 2010, at 1:20 PM, Gavin Henry wrote:

http://wiki.freeswitch.org/wiki/FreeSwitch_FAQ#Q:_How_many_concurrent_calls_...

Hi Brian, OK, then It should clearly say what Anthony said or something like it: "Please do not ask this question on the mailing lists as you will always get the same official response from the FreeSWITCH project; "we only perform benchmarking and confirm these results per FreeSWITCH deployment, as each deployment will result in varying figures. Commercial support is available from the project for this task. The project has learned from experience the dangers of entertaining such questions and its policy is to not do so over the free public forum." Thanks. -- http://www.suretecsystems.com/services/openldap/ http://www.suretectelecom.com

2010/2/13 João Mesquita <jmesquita@freeswitch.org>:

The wiki is public, you know?

Of course I know. I don't represent the project so couldn't obviously add that so it was in anyway official. I'm not trying to be picky here only make things clear in the docs for when the next person comes along and the FreeSWITCH team can just answer an email with the link to the FAQ on the wiki. -- http://www.suretecsystems.com/services/openldap/ http://www.suretectelecom.com

On 13 February 2010 22:07, Peder <peder@networkoblivion.com> wrote:

Instead of complaining about this, why don't you benchmark it yourself on your hardware and post your results on the wiki for others to see in the future?  Anybody can edit and add info to the wiki.

By the way, if you stay on the list for any length of time you will see that about once a week, someone says "I want to do load testing, please tell me how", or "I did a load test and didn't get x calls per second, why".  I'm not even involved in the development of this project and it gets old seeing that same question over and over again.  Really, the standard answer should be "there is no software limit, it depends on your hardware".  If you want to know the limits of your setup, test it and see what it is.

Did you read any of this thread at all? The OP asked about techniques to benchmark his kit, and any rough figures. The projects answer was to hire them for some commercial support. So anything I post on the wiki is not official. My suggestion is merely to put the official project stance on any benchmark questions or post to the wiki some tips on how to run your own tests with a link to getting them verified by the team via the commercial support. In no way am I complaining, just trying to get this cleared up for the next time someone asks. That's the whole point of docs and why I enjoy writing them. Write them once and refer people to them, fixing doc bugs along the way. Gavin. -- http://www.suretecsystems.com/services/openldap/ http://www.suretectelecom.com

One thing that seems to help a lot is to have a "Success Stories" page. A lot of the previous emails emphasize the point of benchmarking FS and hitting it hard with a suite of tools available to do just that. That all makes sense. The problem is that for a lot of newcomers, benchmarking doesn't really make a lot of sense right away. In more details ... the process would be something like this: a) I need to use a system that's able to do certain things for a conference for instance (happens to be my case). It needs to be flexible and do a bunch of things. b) FS seems like the best candidate out there "today". c) Get a whatever server and install FS on it. Nice the thing runs, it seems flexible, good performance, I can't believe my own eyes. Good product. d) Let me build an application around it now. Here is an ESL, WEB API, cool. The sky is the limit. e) Application is now built. f) Let's benchmark ... Ok, let me now buy what I think the best server will be and it'll cost $2000.00. g) FS is amazing, I am getting these many concurrent calls, active sessions, things are great for this X system that I bought. I will scale up and get 5 more of the same type of server. Now the question is, was that the right choice to make? Having access to a place where others post their results help you make a better determination on what kind of platforms to run your system on. Someone with whom I've had a nice exchange of words in the past threw something out there that was really helpful ... "Modern machine with dual core, 64 bit for best results"; if my memory serves me right. This is the kind of information that would be needed to then go out and buy the server/s and then do the actual benchmarking and correctly engineer/architect the platform. If I see that some of the success stories are using Servers X with Foo CPU/s and bar MEM, then that would be for me a good indication of the direction or areas to investigate, go into. If I then want top of the line tuning, excellent benchmarking and many more things, yeah having the commercial support and having to pay for it it's def. not a problem. After all a lot of work and hours go into developing the product and we are getting it free. That part is perfectly understood and supported as well as the development of some more specific features. Alfredo On Sat, Feb 13, 2010 at 5:48 PM, Anthony Minessale <anthony.minessale@gmail.com> wrote:

Helping with docs, now there is a topic we all groove on.

On Feb 13, 2010 4:43 PM, "Gavin Henry" <gavin.henry@gmail.com> wrote:

On 13 February 2010 22:07, Peder <peder@networkoblivion.com> wrote:

...

Instead of complaining about t...

Did you read any of this thread at all? The OP asked about techniques to benchmark his kit, and any rough figures. The projects answer was to hire them for some commercial support. So anything I post on the wiki is not official. My suggestion is merely to put the official project stance on any benchmark questions or post to the wiki some tips on how to run your own tests with a link to getting them verified by the team via the commercial support.

In no way am I complaining, just trying to get this cleared up for the next time someone asks. That's the whole point of docs and why I enjoy writing them. Write them once and refer people to them, fixing doc bugs along the way.

Gavin.

-- http://www.suretecsystems.com/services/openldap/ http://www.suretectelecom.com

_____________...

_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

We have install guides and yet people still have problems installing FreeSWITCH. shrug! More docs are a plus. /b On Feb 13, 2010, at 4:07 PM, Peder wrote:

Instead of complaining about this, why don't you benchmark it yourself on your hardware and post your results on the wiki for others to see in the future? Anybody can edit and add info to the wiki.