Dear Fellows,
The following is no new thought, but a blinding flash of the obvious ... I
suppose we are all in the "requirements collection" phase. Nevertheless it
is a good idea to check what is available right now. This is not really
"implementation", but merely "investigating what has been required at some
point in time". The more input (and reference code, for later) the better.
The next is more important for me: logging and time synchronisation should
be completely independent from each other. While we certainly all do agree
how important synchronisation is (and NTP/xntp is a Good Thing[tm]), the
delivery of logging information should not be prohibited if hosts are out
of sync. And even if timing is required for authentication, it is a matter
of the authentication protocol and not the logging.
>From my own personal experience I can tell you that even in a "clean" NTP
environment with multiple GPS receivers and stratum 2/3/4 servers time is
not as ... common as you would expect.
The easiest way to handle this problem might be to send a timestamp with
the message and record the received and local time at the target machine.
Could well be as a delta. But this is already implementation :-)
Volker
--
Volker Wiegand Phone: +49 (0) 6196 / 50951-24
SuSE Rhein/Main AG Fax: +49 (0) 6196 / 40 96 07
Mergenthalerallee 45-47 Mobile: +49 (0) 179 / 292 66 76
D-65760 Eschborn E-Mail: [EMAIL PROTECTED]
++ Only users lose drugs. Or was it the other way round? ++
