Hello, We've configured a centralized logs server (Solaris 10) in order to collect and manage all log messages coming from +/- 100 servers (Solaris, Linux,...). For an unknown reason, the logs coming from one of our server (Debian) are coming in the following format on the log server : Sep 7 10:08:37 server1 PAM_unix[6142]: authentication failure; (uid=0) -> delphine for ssh service As you can see, we only receive the hostname (server1) but not the FQDN of this server (server1.ourdomain.be). For all the other servers, we have the FQDN in the logs. Here is an example with server2.ourdomain.be (Debian) : Sep 7 10:11:01 server2.ourdomain.be/server2.ourdomain.be PAM_unix[27542]: authentication failure; (uid=0) -> delphine for ssh service The only difference between server1 and the other ones is that it uses syslog-ng instead of syslog in order to send its logs. Here are the options used in the configuration files. 1° On the log server options { create_dirs(yes); dir_perm(0705); dir_owner(root); perm(0600); owner(root); sync(0); check_hostname(no); use_fqdn(yes); use_dns(yes); dns_cache(yes); dns_cache_expire(604800); dns_cache_size(400); stats(60); keep_hostname(yes); chain_hostnames(yes); }; 2° On server1 options { use_fqdn(yes); use_dns(yes); keep_hostname(yes); chain_hostnames(no); long_hostnames(no); sync(0); }; 3° On server2 we are using syslog instead syslog-ng Any idea ? Thanks. _________________________________________________________________ Grand passioné ? Rassemblez tout ce qui vous intéresse en un seul endroit ! http://get.live.com/live/features