Hi people!
After a bit of research and lib installations, I've managed to build the full kamailio deb packages for a raspberry pi 3 (version B) running raspbian stretch (lite) that can be found at [1] (with a really few module exceptions that might be used in really specific scenarios -> I think)
Just for fun, I'm planning to do some CPS stats tomorrow (if I got the time), with a basic config and also share here, but, if someone else tried it before and is able to share some stats/benchmarks, would be greatly appreciated (and interesting to compare, given my "decent" rpi overclock), because I'm also curious about VoIP stuff on raspberry pi! :D
[1] https://github.com/smititelu/rpi/tree/master/kamailio/5.1.1
Thank you, Stefan
Hello,
I haven't played with kamailio on raspberry pi 3, I did it with an older release of pi, however I haven't stress test it.
But I did install Kamailio on Ubuntu Phone and then used sipp to generate some traffic (not preparing for production, but for a dangerous demo). What I noticed that as soon as kamailio started to use too much cpu for long time, it was killed. I didn't invest much time to dig in, but my guess was that there was some energy protection system that reacted.
I thought it's an interesting aspect that should be considered when running on systems relying on batteries (e.g., for IoT).
Cheers, Daniel
On 06.02.18 23:40, Mititelu Stefan wrote:
Kamailio running on rpi 3.
All the PI specs, Kamailio basic configs and SIPp xml scenarios I used can be found at [1]. All the tests were done in a LAN, using PI's ethernet interface, running kamailio 5.1.1. All the tests ran for 60 seconds, a couple of times.
1. REGISTER/200, __with db_text__: - at 900 cps test did finish: all UAC registered; pi htop threads were ~15-20% - at 950 cps test did NOT finish: got "Overload warning" on my UAC/UAS SIPp testing machine :(
2. INVITE/180/200/PAUSE(3sec)/BYE/200, __with no media__: - at 370 cps test did finish: all UAC->UAS calls completed; ~150 "180 Trying" Unexpected-Msg on UAC side; pi htop threads were ~50% - at 380 cps test did NOT finish: few(~5) UAC->UAS calls not completed; pi htop threads were ~50%
The results are really impressive !!! (even if the used testing configs were really basic ones). Moreover, I think that I've reached the limit of my current SIPp testing machine, but not of PI's...
Interesting, thanks for sharing, likely I will post a news on kamailio.org about these results, might be useful for others in the future.
Cheers, Daniel
On 14.02.18 16:15, Mititelu Stefan wrote:
-
Daniel-Constantin Mierla -
Mititelu Stefan