Hi, Agalya!
Thank you for including the POST method in your tests as well. In order
to understand more about the problem, we need to print some information,
in order to correlate the SIP flow of those calls with the fact that
resume_route was not called, although the HTTP requests are correctly
handled. Could you please do the following info to your testing
framework, and attempt to replicate the missing resume_routes again?
- trace all traffic (pcap format) on your server's SIP port (e.g. 5060).
Example command: "tcpdump -i <interface> -s 5060 -w <some-file>"
- in your OpenSIPS script:
* for each INVITE, print the $civariable
* for each resume_route, print the $ci variable again
If, by any chance, all theabove information is already available, could
youplease email the pcap and problematic Call-IDs to [email protected],
so I can take a look? Thanks!
Best regards,
Liviu Chircu
OpenSIPS Developer
http://www.opensips-solutions.com
On 17.10.2016 18:33, Ramachandran, Agalya (Contractor) wrote:
Hi Razvan/Liviu,
My test observance as follows:
REST_PUT:
1)In the sipp client if I have pause value after 200 OK between 12 to
24 sec, am seeing that for ~15 calls resume_route is not called.
2)I have increased pause time to 24 -45 sec and also 180-240 sec.
This both cases, call setup is complete. resume_route is executed for
all 100K calls.
REST_POST:
3)In the sipp client, pause of 12 to 24 sec after 200OK. For 10 calls
out of 100K calls, resume_route is not called.
This happened only once out of three times. Other 2 times call set is
complete. REST_POST is returning 4xx in my case.
4)In the sipp client, pause of 24 to 45 sec after 200OK.Call set up is
complete.
Regards,
Agalya
*From:*[email protected]
[mailto:[email protected]] *On Behalf Of *Razvan Crainea
*Sent:* Thursday, October 13, 2016 10:16 AM
*To:* [email protected]
*Subject:* Re: [OpenSIPS-Users] Pending OpenSIPS minor releases: Last
minute bug fixes!
Hi, Agalya!
Can you try running the tests with the stock code, without the PUT API
code? Just run the tests with the POST method, to confirm whether
there is a bug in the stock code, or in the custom code you made.
Best regards,
Răzvan Crainea
OpenSIPS Solutions
www.opensips-solutions.com <http://www.opensips-solutions.com>
On 10/13/2016 04:55 PM, Ramachandran, Agalya (Contractor) wrote:
Hi Liviu,
Thanks for sharing the process for source code contribution.
Please find my response inline.
Just to be sure: we're talking about REST POST now, right?
Agalya: No. Am using REST PUT API.
Regarding the issue: are there any OpenSIPS log errors that might
help us?
Agalya: No errors in the logs.
Also, did you deduce the 99986 number by grepping the logs, or by
looking at the completed SIP calls? I'd recommend the latter,
since log lines may be rate-limited / overlapped, etc.
Agalya: By grepping logs as well by looking completed SIP calls. I
did this for 50,000 calls and the number matched for SIP logs and
call completed in sipp.
In Wireshark, I see response for all 50,000 HTTP requests.
I couldn’t get complete logs for 100K calls may be because as you
said rate-limited / overlapped, etc.
Regards,
Agalya
*From:*Liviu Chircu [mailto:[email protected]]
*Sent:* Thursday, October 13, 2016 4:10 AM
*To:* Ramachandran, Agalya (Contractor)
<[email protected]>
<mailto:[email protected]>; OpenSIPS users mailling
list <[email protected]> <mailto:[email protected]>
*Subject:* Re: [OpenSIPS-Users] Pending OpenSIPS minor releases:
Last minute bug fixes!
Hi, Agalya!
If I get green signal from my management, I will contribute
code for REST_PUT. Can you share me the process to contribute
code ?
First, you create a GitHub account. Then you fork the OpenSIPS
repo [1] to your account. This allows you to work on it
independently and push changes back to GitHub when you're done, so
they are visible for everyone.
Once the fork is done, you clone the forked project on your
machine so you can work on its code. You then apply your custom
patch(es), make the necessary commits and push these changes back
to GitHub.
The process of proposing the merge of a forked project back into
the main project is called a "Pull Request" [2]. This is the final
step of contributing code, and you can easily do it with a few
clicks via GitHub's web interface.
Out of 2 times, I tested I observed the below issue for once.
Before I used to have it for every test.
1.Tried to load 100,000 calls - But route[resume_http] is
called only for 99986 calls.
Every time approximately 10-20 calls, route[resume_http] is
not called. But if I see the tcpdump, I am seeing 100,000 HTTP
request and 100,000 HTTP 200 OK responses.
When printing the response in resume_http for every call-id,
10-20 calls response is not printed - which means resume is
not called for these calls.
Am not filtering any response code.
Any clue on this one?
Just to be sure: we're talking about REST POST now, right?
Regarding the issue: are there any OpenSIPS log errors that might
help us? Also, did you deduce the 99986 number by grepping the
logs, or by looking at the completed SIP calls? I'd recommend the
latter, since log lines may be rate-limited / overlapped, etc.
[1]: https://github.com/OpenSIPS/opensips
[2]: https://github.com/OpenSIPS/opensips/pulls
_______________________________________________
Users mailing list
[email protected] <mailto:[email protected]>
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
_______________________________________________
Users mailing list
[email protected]
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
_______________________________________________
Users mailing list
[email protected]
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
