Hello,
Ok, I will check that.
In terms of call processing no matter how many non-numeric entries mtree
has?
Thank you for you help
BR
José
2016-07-01 12:59 GMT+01:00 Daniel-Constantin Mierla :
> Hello,
>
> well, it is not easy to predict, because it is a matter of density (or
> overlapping ch
Hello,
well, it is not easy to predict, because it is a matter of density (or
overlapping characters).
Anyhow, it also depends on the module parameter char_list -- if you want
any character from alphabet there is going to use lot of memory.
For first value added in the tree, it creates 'sizeof(v
Hello,
Can you tell me what is the memory usage in mtree for each non-numeric
record with ~ 10 characters?
Mtree using more memory can decrease its own performance?
Thank you
BR
José
2016-07-01 12:34 GMT+01:00 Daniel-Constantin Mierla :
> Hello,
>
> the matching rules are stored pre-compiled i
Hello,
the matching rules are stored pre-compiled in memory, not being parsed
every time, unless you have rules with variables.
Combining mtree with dialplan can increase performances, just be aware
that if you have non-numeric values stored in 'tprefix' column, it is
going to use a lot of memory
Hello Daniel,
Thank you for your reply,
What are optimizations to increase the speed of processing?
I have thought in another solution that is use diaplan module only for
apply manipulations and use mtree to match prefixes after apply
manipulations.
What is your opinion on this?
Thank you again
Hello,
the transalation records are cached by kamailio and precompiled, so
there are optimizations to increase the speed of processing. But have in
mind that the rules are matched one by one from the same dpid, in the
order of priority.
For the records where you don't have substitutions/replaceme
Hello there,
I'm evaluating if I can use dial-plan and dispatcher module to build a
kamailio routing server to route calls based on number dialed(sometimes the
number dialed can be Alphanumeric).
I'm expecting have lot of dial-plan entries, some of them with
manipulations, others without.
Anyone c
