Hi List,
Anybody using this in production yet? If so what kind of volume and how is reliability?
Looking to move to this platform, looks very good, interested to hear some experiences.
Thanks, Stephen
Using it in several production environments; awesome reliability!
Wouldn't go back to 1.5.x for the world.
-- Sent from mobile device
On Mar 24, 2010, at 4:09 AM, dotnetdub dotnetdub@gmail.com wrote:
On 3/24/10 11:10 AM, Alex Balashov wrote:
Using it in several production environments; awesome reliability!
Thanks Alex! And you know, this is not by chance...
During last month I have been traveling a lot, I met people that use 3.0 in production or they passed the internal testing phase, being just to switch it to production. They were pleasantly surprised by the results.
Since this question pops up from time to time, I tried to collect the facts and procedures during past year and demystify why 3.0.x is very stable. Probably I forgot to mention other people or companies that substantially contributed to 3.0.x (drop me an email to fix it) ... anyhow, here is the link:
http://www.kamailio.org/w/2010/03/remarks-about-v3-0-x-strong-stability/
Cheers, Daniel
Hi Daniel,
thanks for the update, but seems to be that many users wait for 3.1 which was announced as the first real sip-router release and is again a major change which need effort to validate.
Is there already a target date proposed for 3.1?
regards, Andreas
2010/3/31 Daniel-Constantin Mierla miconda@gmail.com
On Wednesday 31 March 2010, Andreas Heise wrote:
Hello Andreas,
well, 3.1 will have an even better integration, to further blurry the border between the different parts coming from ser and kamailio. So in my opinion, if people wait, its more because of a 'don't trust a .0 release..' opinion. ;-)
Is there already a target date proposed for 3.1?
I don't think so, Daniel please correct me if i'm wrong. Normally it should be out something roughly six monts after 3.0, this would mean somewhere in July.
Regards,
Henning
Hi Henning,
On 3/31/10 6:17 PM, Henning Westerholt wrote:
indeed, no fixed date by now. We should schedule a irc devel meeting soon to sketch the roadmap for 3.1. There is a lot of new features, some still need more polishing to be completed (at least what i started) and the other work is related to merging some duplicated modules.
In another email I said maybe summer will be used for testing and release in beginning of autumn so we have approx 8 months cycle.
Cheers, Daniel
Hi Andreas,
On 3/31/10 4:29 PM, Andreas Heise wrote:
I think it was a mistake in communication with the achievements for 3.0 -- we have to work on and fixed it.
When we met in Karlsruhe, Nov 2008, we kind of agreed as plan go on to make the core and tm work with kamailio modules or ser modules. The option would have been either at compile time or at startup.
However, we succeeded to go one step further and actually be able to run mixed modules at the same time. So we are now, considering the initial goals, pretty much the 3.1:
- no need to compile with different flags to run one or the other type of modules - no need to configure what type of modules are running
Simply you can mix them, it works.
I agree that for 3.1 we will be better from integration point of view, in regards to less duplicated modules (hopefully sl, domain, pdt, ... will be merged). However some of the modules will stay as they are now for longer time. Here are auth_db, usrloc, and the other modules that differ as backend database structure (e.g., user profiles tables, location). There are public (siremis, serweb) and private tools, povisioning and monitoring systems that cannot be dropped easily.
So we will have flavours packaging for a while. Otherwise, 3.1 won't have other integration work for core and tm, that work is finished. What comes in those parts of code for 3.1 are pure brand new features.
Kamailio 3.0 just enables some features by default, sets different default behavior which can be tuned by parameters anyhow. Here I tried to collect more details lately: http://sip-router.org/releases/ http://sip-router.org/kamailio-release/
There is no patch that has to be applied in kamailio 3.0 branch order to compile ser flavour out of it.
Is there already a target date proposed for 3.1?
Not clearly decided, but the usual 6-8 months is still in place, that means testing should start beginning of summer, which will result in release maybe beginning of autumn (not to do it from the beach :-) in vacation).
Cheers, Daniel
-
Alex Balashov -
Andreas Heise -
Daniel-Constantin Mierla -
dotnetdub -
Henning Westerholt