[syslog-ng]TCP transport reliability

[Balazs Scheidler]

> I'm testing syslog-ng with TCP transport, and I'm a little surprised-
> perhaps I'm doing something wrong? I'm finding severe message loss,
> which I had imagined TCP would eliminate. Briefly, my testing paradigm is
> as follows:

What are you using to send log messages? shell script with logger, or a
custom program or? can you post its source?

> lwp_sema_post(0xFF00FE78)                       = 0
> lwp_sema_wait(0xFF00FE78)                       = 0
> lwp_mutex_wakeup(0xFF150C58)                    = 0
> lwp_mutex_lock(0xFF150C58)                      = 0
>     Received signal #15, SIGTERM [default]
>       siginfo: SIGTERM pid=2539 uid=0
> lwp_sigredirect(1, SIGTERM)                     = 0
>         *** process killed ***
> 
> 
> Does anyone know offhand why this should happen? 
> 
> Thanks very much for any advice you can offer.

syslog-ng forks when started but waits for its child to signal whether
initializing the configuration was successful. When the config is
initialized the child sends a SIGTERM is sent to its parent, exactly the
signal it was waiting for.

-- 
Bazsi
PGP info: KeyID 9AF8D0A9 Fingerprint CD27 CFB0 802C 0944 9CFD 804E C82C 8EB1
     url: http://www.balabit.hu/pgpkey.txt