Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Issues with RTPProxy process #125

Open
ergyspapa opened this issue Feb 27, 2022 · 7 comments
Open

Issues with RTPProxy process #125

ergyspapa opened this issue Feb 27, 2022 · 7 comments
Labels

Comments

@ergyspapa
Copy link

Dear team,
I have installed rtpproxy on Centos7 and I am using it with Sippy B2BUA as proxy. I am currently facing some issues as during calls I am getting No Online Rtp Proxy is found.
I checked the rtpproxy process and it is running without any issues. I also activated debug to check for any errors but can find anything as no error is showing.
I am currently testing it with 500cps and 5000 concurrent calls to check the full load.
Rtpproxy and Sippy B2BUA are running in different dedicated machines.
I have also run rtpproxy with root user and ulimit is rised at 40000.
Also need to mention that after some minutes of high load the calls are stopped and then they start again for a few minutes and they go down again.
The only doubt that I see is that the process of rtpproxy is not going over 100% on single core.
I would really appreciate some help or tips that what I am doing wrong.

Rtpproxy version is: 2.2 from git.
Udp ports are: 5000:50000
No firewall is running during tests.
Servers are both: E5 - dedicated with ssd and 64Gb or Ram.

@sobomax
Copy link
Member

sobomax commented Feb 3, 2023

Ergys, for the target amount of the traffic you'd need 4-5 rtpproxy instances running in parallel to handle properly. Our rule of thumb is about 1,000-1,500 of sessions per instance. Each instance is able to utilize ~2-3 cores, so this would ensure the load is well balanced across the machine. If the RTPProxy would be controlled by Kamailio then there is already feature allowing you to load balance across multiple instances by listing individual sockets in the configuration parameters, in case if you want to use B2BUA then there is go-rtp_cluster middleware module allowing to do the same.

@mrsanvicente
Copy link

Hello @sobomax / all, what´s the recommended way to setup multiple rtpproxys?
I have configured multiple but they point to the same binary rtpprxy. and at some point all rtpproxy services fails at once.

Thank you

@sobomax
Copy link
Member

sobomax commented Mar 6, 2024

hey @mrsanvicente there are plenty ways to do it. You can create bunch of .service files with slightly varying parameters (i.e. rtpproxy.1.service, rtpproxy.2.service, etc etc). Or you just create a wrapper script that starts takes common arguments and starts many copies under the hood and hook it up to your existing startup script. You should be careful they don't clash over the same resource (i.e. TCP or UDP port or socket name) but other than that should work.

@sobomax
Copy link
Member

sobomax commented Mar 7, 2024 via email

@sobomax
Copy link
Member

sobomax commented Mar 7, 2024

@mrsanvicente we've been running such configuration in production since 2016-2017, doing 20k+ RTP sessions across 5 nodes, each running 4 RTP proxies.
ScreenShot267

@mrsanvicente
Copy link

mrsanvicente commented Mar 7, 2024 via email

@mrsanvicente
Copy link

mrsanvicente commented Mar 8, 2024 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

3 participants