testing about bbb.

[allen]

Hi, all

 Recently,we carried out some basic tests to big blue button v0.8.

  Test environment:

  OS: Ubuntu 10.04.4 LTS

Hardware: Linux 2.6.32.49 2012 x86_64

CPU info = {

Processor: 5

vendor_id: GenuineIntel

CPU family: 6

model: 45

model name: Intel (R) Xeon (R) CPU E5-2620 0...@2.00GHz

stepping: 7

cpu MHz: 2000.099

cache size: 15360 KB

fpu: yes

fpu_exception: yes

cpuid level: 10

wp: yes

flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush mmx fxsr sse sse2 ss syscall nx lm constant_tsc rep_good pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt aes avx hypervisor lahf_lm

bogomips: 4000.43

clflush size: 64

cache_alignment: 64

address sizes: 46 bits physical, 48 bits virtual

} * 6

Mem total: 16G

Network card: 1000 M LAN

Testing environment:

RTMP rotocol use the product of juv. Reference：http://www.smaxe.com/juvclient.jsf

 Testing Procedure:

  step 1, the test script registered five Share Objects,it's chatSO, participantsSO, presentationSO, drawSO, recordControlSO.

  

  step 2, with the "student" side of the two machines simulation, and start 10 processes, each process start 50 threads on the another testing machine.

  

  step 3, every time to start one process monitoring and meantimethe red5 testing machine the index.

  

  Getting different results in different versions:

  

Red5 v1.0.0.4389 version

step-by-step to start 20 processes, when the thread close to 800, the teachers client use the painting brush to finish some actions,which is most dissipative resource, the testing machine cpu usage instantaneous more than 90%, the memory achieve to 500M, NIC occupancy 25M or so.

In this case, the student client machine is unresponsive to teacher's action.

Upgrade to Red5 v1.0.0.4638

This red5 version has slowed reaction  only connection the server with 300 threads.

  

  This test use the distinct red5 version,and big blue button app is same. So,what's reason lead to these difference? and main reason is the side of the red5 Or bbb?

Expect your answer, thank you.

[Fred Dixon]

Hi Allen,

First off -- many thanks for posting the results of this test.  We'd be interested if you could run the same tests with the latest build of BigBlueButton 0.81-dev.  See

   https://code.google.com/p/bigbluebutton/wiki/081InstallationUbuntu

We're using the latest builds of red5 (as of the date of this post) in 0.81-dev and would be curious if you find any differences in your tests.

Regards,... Fred

-- 
BigBlueButton Developer

http://bigbluebutton.org/
http://code.google.com/p/bigbluebutton

BigBlueButton on twitter: @bigbluebutton

--
You received this message because you are subscribed to the Google Groups "BigBlueButton-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bigbluebutton-...@googlegroups.com.
To post to this group, send email to bigblueb...@googlegroups.com.
Visit this group at http://groups.google.com/group/bigbluebutton-dev?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[HostBBB.com]

Allen, as you scale, some of the OS, java, and tomcat limits get
reached.

What changes to ulimits, java options, and tomcat configs did you do
for your testing? Is this with stock Ubuntu and BBB install settings?

Would be interested, we have had to tweak these to get above 1024
streams in our testing.



regards.
Stephen
hostbbb.com

On May 15, 8:17 am, Fred Dixon <ffdi...@gmail.com> wrote:
> Hi Allen,
>
> First off -- many thanks for posting the results of this test.  We'd be
> interested if you could run the same tests with the latest build of
> BigBlueButton 0.81-dev.  See
>
>    https://code.google.com/p/bigbluebutton/wiki/081InstallationUbuntu
>
> We're using the latest builds of red5 (as of the date of this post) in
> 0.81-dev and would be curious if you find any differences in your tests.
>
> Regards,... Fred
> --

> BigBlueButton Developerhttp://bigbluebutton.org/http://code.google.com/p/bigbluebutton

> BigBlueButton on twitter: @bigbluebutton
>
>
>
>
>
>
>
> On Mon, May 6, 2013 at 9:33 PM, allen <zoyuc...@gmail.com> wrote:
> > Hi, all
>
> >  Recently,we carried out some basic tests to big blue button v0.8.

> > *  Test environment:*

> > *Testing environment:*

>
> > RTMP rotocol use the product of juv. Reference：
> >http://www.smaxe.com/juvclient.jsf
>

> > * Testing Procedure:*

>
> >   step 1, the test script registered five Share Objects,it's chatSO,
> > participantsSO, presentationSO, drawSO, recordControlSO.
>
> >   step 2, with the "student" side of the two machines simulation, and
> > start 10 processes, each process start 50 threads on the another testing
> > machine.
>
> >   step 3, every time to start one process monitoring and meantimethe red5
> > testing machine the index.
>
> >   Getting different results in different versions:
>

> > *Red5 v1.0.0.4389 version*

>
> > step-by-step to start 20 processes, when the thread close to 800, the
> > teachers client use the painting brush to finish some actions,which is most
> > dissipative resource, the testing machine cpu usage instantaneous more than
> > 90%, the memory achieve to 500M, NIC occupancy 25M or so.
>
> > In this case, the student client machine is unresponsive to teacher's
> > action.
>

> > *Upgrade to Red5 v1.0.0.4638*
>
> > This red5 version has slowed reaction  only connection the server with* 300
> > threads*.
>
> >  * This test use the distinct red5 version,and big blue button app is

> > same. So,what's reason lead to these difference? and main reason is the

> > side of the red5 Or bbb?*

[Fred Dixon]

HI Allen,

In stress testing red5, you will likely need to give red5 more file descriptors.  You can set

  ulimit -n 65535

at the top of the /etc/init.d/red5 script, then restart red5.  This setting will increase both the soft and hard limit for the number of file descriptors allowed for red5 from 1024 to 65535.

Regards,... Fred

-- 
BigBlueButton Developer

http://bigbluebutton.org/
http://code.google.com/p/bigbluebutton

BigBlueButton on twitter: @bigbluebutton