Timo,

    Thanks for this reference - I actually have been playing / tweaking with various mentioned items in this reference earlier.

My current settings are as follows:

connect_timeout = 30
net_read_timeout = 180 
net_write_timeout = 180 

The issue still seems to be quite prevalent.  I will say that the amount of aborted connects currently (41) is way lower than the actual amount of syslog lost connection errors.

Sincerely,
Brandon Armstead

On Thu, Dec 8, 2011 at 2:48 PM, Timo Reimann <sr@foo-lounge.de> wrote:
Hey Brandon,


Am 08.12.2011 um 20:32 schrieb Brandon Armstead:
>     I am very very very frequently seeing in logs "Lost connection to MySQL Server" --- the MySQL server that it is connecting to is less than 1 ms away (essentially sitting in the same rack).
>
> Dec  8 19:21:15  /usr/local/sbin/kamailio[4149]: INFO: dialog [dlg_db_handler.c:584]: trying to remove dialog [c9d5ed0-13c4-4ee10778-92e3eb21-a90244c@XXX.XXX.XXX.XXX], update_flag is 4
> Dec  8 19:21:16  /usr/local/sbin/kamailio[4149]: ERROR: db_mysql [km_dbase.c:122]: driver error on query: Lost connection to MySQL server during query
> Dec  8 19:21:16  /usr/local/sbin/kamailio[4149]: ERROR: <core> [db_query.c:242]: error while submitting query
> Dec  8 19:21:16  /usr/local/sbin/kamailio[4149]: ERROR: dialog [dlg_db_handler.c:598]: failed to delete database information
>
> *** MOD PARAM CONFIG ***
>
> # database ping
> modparam("db_mysql", "ping_interval", 150)
> modparam("db_mysql", "auto_reconnect", 1)
>
> Any thoughts on what may be happening?  It seems the overall connection (network) between the two servers is great.

The MySQL documentation enumerates a number of reasons:

http://dev.mysql.com/doc/refman/5.1/en/error-lost-connection.html
(This is the URL to MySQL version 5.1 documentation.)

You may want to check those for possible matchings.


Cheers,

--Timo