[SR-Users] Which algorithm to use as load-balance

[David Villasmil]

Hello Daniel,

I compiled and i'm testing 5.2, but it doesn't seem to be doing what I was
expecting, my config is as follows:

route[DISPATCH] {
    dp_match("1", "$ru");

    if(!ds_select_routes("1=4;11=9", "ruri", "10"))
    {
        send_reply("404", "No destination");
        exit;
    } else {
        xlog("[DISPATCH] ds_select_dst was succesful\n");
    }
    xlog("L_DBG", "--- SCRIPT: going to <$ru> via <$du>\n");
    t_on_failure("RTF_DISPATCH");
    route(RELAY);
    exit;
}

And my table is:

+----+-------+------------------------------------+-------+----------+----------------------------+-------------+
| id | setid | destination                        | flags | priority |
attrs                      | description |
+----+-------+------------------------------------+-------+----------+----------------------------+-------------+
| 213 |     1 | sip:34.201.28.49:5080              |     0 |       10 |
weight=100                | node1-proxy |
| 229 |     1 | sip:54.209.229.198:5080            |     0 |       10 |
weight=100                | node2-proxy |
| 232 |    11 | sip:35.170.67.206:5080             |     0 |       10 |
weight=100                | node3-proxy |
+----+-------+------------------------------------+-------+----------+----------------------------+-------------+

But kamailio is always sending to the same ip 198, although it should be
load-balancing round-robin to the 2 nodes on setid 1. And when those 2
nodes have 10 calls, then start sending to setid 11.

Is my config wrong?

Regards,

David Villasmil
email: david.villasmil.work at gmail.com
phone: +34669448337

ᐧ

On Thu, Nov 15, 2018 at 4:09 PM David Villasmil <
david.villasmil.work at gmail.com> wrote:

> Ah, 5.2 branch. Will test it, thanks!
>
> On Thu, 15 Nov 2018 at 16:06, Daniel-Constantin Mierla <miconda at gmail.com>
> wrote:
>
>> Hello,
>>
>> besides master, you can also use branch 5.2 -- this is going to be used
>> to release 5.2.0 and other future 5.2.x versions. Those two branches have
>> the same version of the module at this moment.
>>
>> Cheers,
>> Daniel
>> On 15.11.18 16:48, David Villasmil wrote:
>>
>> Hey Daniel,
>>
>> Thanks for replying. Can I test this from master?
>>
>> David
>> On Thu, 15 Nov 2018 at 15:12, Daniel-Constantin Mierla <miconda at gmail.com>
>> wrote:
>>
>>> Hello,
>>>
>>>
>>> On 14.11.18 23:16, David Villasmil wrote:
>>> > Hello,
>>> >
>>> > Is it possible to set an algorithm like the following?
>>> >
>>> > fs1 - Priority 10 - maxload 20 calls
>>> > fs2 - Priority 10 - maxload 20 calls
>>> > fs3 - Priority 9 - maxload 20 calls
>>> > fs4 - Priority 8 - maxload 20 calls
>>> > fs5 - Priority 7 - maxload 20 calls
>>> >
>>> > So that, whenever any new call comes in, fs1 and fs2 would be used,
>>> > meaning the fs with the highest priority would be use UNLESS there's
>>> > 20 calls on that server, in which case, the next server with the
>>> > highest priority would be used.
>>> >
>>> this should be easier possible with the upcoming 5.2.0 (to be released
>>> in about 2 weeks or so), see:
>>>
>>>   -
>>>
>>> https://www.kamailio.org/docs/modules/devel/modules/dispatcher.html#dispatcher.f.ds_select_routes
>>>
>>> For older versions you may need to do more complex logic in the
>>> configuration file to call dispatcher for different groups at different
>>> stages of failure re-routing.
>>>
>>> Cheers,
>>> Daniel
>>>
>>> --
>>> Daniel-Constantin Mierla -- www.asipto.com
>>> www.twitter.com/miconda -- www.linkedin.com/in/miconda
>>> Kamailio World Conference -- www.kamailioworld.com
>>> Kamailio Advanced Training, Nov 12-14, 2018, in Berlin -- www.asipto.com
>>>
>>> --
>> Daniel-Constantin Mierla -- www.asipto.comwww.twitter.com/miconda -- www.linkedin.com/in/miconda
>> Kamailio World Conference -- www.kamailioworld.com
>> Kamailio Advanced Training, Nov 12-14, 2018, in Berlin -- www.asipto.com
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.kamailio.org/pipermail/sr-users/attachments/20181116/6888f519/attachment.html>