Hi Juha,
Thanks for your reply.
First prominent doubt that I have is; when I am using LCR in db_only mode why does openser try to load the table while starting up?
The scenario is as follows:
I had 50 rows in LCR table and started openser and it started properly and also routed the calls properly according to the values in lcr and gw table. Now I just appended more 40,000 rows to this table and again the routing worked properly
as per my lcr table. This behaviour was correct as I was using query only mode. I did not need any restart or any fifo reload command.
Only when I restarted my openser, the openser again died as it was trying load the lcr table while starting.
Could this be a bug in lcr module as I am not sure whats happening.
I tried increasing MAX_NO_OF_LCRS but even that did not help.
Somebody please help if poosible.

Thanks,

w/regards,
jayesh

----- Original Message ----
From: Juha Heinanen <jh@tutpro.com>
To: Jayesh Nambiar <voip_freak@yahoo.co.in>
Cc: openser <users@openser.org>
Sent: Tuesday, 26 December, 2006 10:58:01 PM
Subject: [Users] LCR db_only mode problem

X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil]
the error messages you get are coming from this statement:

    if (lcr_dbf.query(dbh, NULL, 0, NULL, lcr_cols, 0, 4, 0, &res) < 0) {
        LOG(L_ERR, "lcr_reload_gws(): Failed to query lcr data\n");
        lcr_dbf.close(dbh);
        return -1;
    }

i don't know which kind of memory db queries reserve.  if shared memory,
you can start openser with larger -m parameter value.  if pkg memory,
you need to edit PKG_MEM_POOL_SIZE in config.h and recompile openser.

with lots of lcr entries and no caching, performance may be bad.
if you use caching then you need to increase value of

#define MAX_NO_OF_LCRS 256

in lcr_mod.c and recompile lcr module.  increasing MAX_NO_OF_LCRS means
that more shared memory will be used so you may also need to start
openser with larger -m parameter value.

-- juha