Hi, If you enable remote syslog (@tls4://log.example.com:6514) in /etc/syslog.conf and syslogd is started before network is available then the machine will never log to remote syslog server even after network is up.
This is due to failure in name resolution which is permanent (no retries). This happens because the name resolution is in config file checks cfline() [priv]: msg PRIV_GETADDRINFO received syslogd: bad hostname "@tls4://log.example.com:6514" If you reload the syslogd server, after network is ready then logging will start. If you add the IP of the remote server in /etc/hosts then you get "Network is down" and syslogd will retry until is connected which is logical. syslogd: connect "@tls4://log.example.com:6514": Network is down logmsg: pri 053, flags 0x4, from dev, msg syslogd: connect "@tls4://log.example.com:6514": Network is down Logging to FORWTLS @tls4://log.example.com:6514 Logging to FILE /var/log/messages tcp connect retry: wait 1 syslogd: connect "@tls4://log.example.com:6514": Network is down logmsg: pri 053, flags 0x4, from dev, msg syslogd: connect "@tls4://log.example.com:6514": Network is down But adding the IP in /etc/hosts is not ideal, nor reloading the syslogd after a while. Sorry I cannot provide a patch for this refactoring, just report it. regards, G
