Hi,
Is there any way to reload the content of the aliases table without restarting Kamailio?
I have this in my cfg file
if (lookup("location")) { xlog("L_WARN", "Callee is online - - M=$rm RURI=$ru F=$fu T=$tu IP=$si ID=$ci\n"); route(TO_USER_SIPCHECKS); route(RELAY); } else if (lookup("aliases")) { xlog("L_WARN", "Callee is in aliases - - M=$rm RURI=$ru F=$fu T=$tu IP=$si ID=$ci\n"); route(TO_USER_SIPCHECKS); route(RELAY); } else { xlog("L_WARN", "Callee is offline - - M=$rm RURI=$ru F=$fu T=$tu IP=$si ID=$ci\n"); t_newtran(); t_reply("486", "Busy here"); exit; }
And changes in the aliases table does is ignored until I restart Kamailio.
Please see:
http://www.kamailio.org/docs/modules/3.1.x/modules_k/usrloc.html#id2995804
On Tue, Jul 19, 2011 at 11:56 AM, Alex Balashov abalashov@evaristesys.com wrote:
Please see:
http://www.kamailio.org/docs/modules/3.1.x/modules_k/usrloc.html#id2995804
Does this means that I have to use db_mode=3 in order to for the changes in aliases to apear without a restart? That would degrade the performance much with the location table lookup.
On 07/19/2011 07:13 AM, Morten Isaksen wrote:
Does this means that I have to use db_mode=3 in order to for the changes in aliases to apear without a restart?
Since there does not appear to be an MI or RPC command in usrloc to reload the table from the database side, yes.
The general assumption of usrloc is that updates are made from within Kamailio and synced to the database for persistence, not the other way around. Perhaps the aliases table is not the optimal mechanism for you.
That would degrade the performance much with the location table lookup.
What makes you say that? How many lookup requests are you throwing at this thing?
Is it more computationally expensive? Yes, absolutely. But unless you're doing tens of thousands of lookups per second, it's inconsequential on contemporary hardware. Also, RDBMs themselves do excellent caching, especially if you give them enough RAM. As Donald Knuth said, "premature optimisation is the root of all evil".
If I'm wrong about how big of a deal this really is, consider making creative use of 'htable' or 'memcache' as a caching strategy.
Hi,
I just learned that you can update the aliases table with "kamctl alias rm" and "kamctl alias add". That solves my problem and I am still able to keep db_mode=2.
But thanks for your help. :)
/Morten
On Tue, Jul 19, 2011 at 1:24 PM, Alex Balashov abalashov@evaristesys.com wrote:
On 07/19/2011 07:13 AM, Morten Isaksen wrote:
Does this means that I have to use db_mode=3 in order to for the changes in aliases to apear without a restart?
Since there does not appear to be an MI or RPC command in usrloc to reload the table from the database side, yes.
The general assumption of usrloc is that updates are made from within Kamailio and synced to the database for persistence, not the other way around. Perhaps the aliases table is not the optimal mechanism for you.
That would degrade the performance much with the location table lookup.
What makes you say that? How many lookup requests are you throwing at this thing?
Is it more computationally expensive? Yes, absolutely. But unless you're doing tens of thousands of lookups per second, it's inconsequential on contemporary hardware. Also, RDBMs themselves do excellent caching, especially if you give them enough RAM. As Donald Knuth said, "premature optimisation is the root of all evil".
If I'm wrong about how big of a deal this really is, consider making creative use of 'htable' or 'memcache' as a caching strategy.
-- Alex Balashov - Principal Evariste Systems LLC 260 Peachtree Street NW Suite 2200 Atlanta, GA 30303 Tel: +1-678-954-0670 Fax: +1-404-961-1892 Web: http://www.evaristesys.com/
SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
On 07/19/2011 07:32 AM, Morten Isaksen wrote:
I just learned that you can update the aliases table with "kamctl alias rm" and "kamctl alias add". That solves my problem and I am still able to keep db_mode=2.
Well, yes. :) I suppose I did not infer from the formulation of your question that this was an acceptable option.
-
Alex Balashov -
Morten Isaksen