On a Solaris 10 b72 machine, I have lines like this in syslog.conf:

  daemon,local5.info              ifdef(`LOGHOST', /var/adm/tcpd.log, @loghost)

Loghost is defined in DNS and the /etc/hosts file, but syslog still
logs only to the local file.  Has something changed since Solaris 9,
or is something mysterious wrong?

-- 
-Gary Mills-    -Unix Support-    -U of M Academic Computing and Networking-

Gary Mills <mills@cc.umanitoba.ca> wrote:
: On a Solaris 10 b72 machine, I have lines like this in syslog.conf:

:   daemon,local5.info              ifdef(`LOGHOST', /var/adm/tcpd.log, @loghost)

: Loghost is defined in DNS and the /etc/hosts file, but syslog still
: logs only to the local file.  Has something changed since Solaris 9,
: or is something mysterious wrong?

Maybe worthless suggestion but check to make sure loghost isn't defined
twice in /etc/hosts. I ran into the same problem playing around with 10,
local file worked, loghost didn't and that was the problem.

I don't remember earlier versions automatically adding loghost to /etc/hosts
defining the machine itself, but it's been a while. 10 seems to do it
automatically.

-bruce
bje@ripco.com

In <cttd5s$hp2$1@e250.ripco.com> Bruce Esquibel <bje@e4500.ripco.com> writes:

>Gary Mills <mills@cc.umanitoba.ca> wrote:
>: On a Solaris 10 b72 machine, I have lines like this in syslog.conf:

>:   daemon,local5.info              ifdef(`LOGHOST', /var/adm/tcpd.log, @loghost)

>: Loghost is defined in DNS and the /etc/hosts file, but syslog still
>: logs only to the local file.  Has something changed since Solaris 9,
>: or is something mysterious wrong?

>Maybe worthless suggestion but check to make sure loghost isn't defined
>twice in /etc/hosts. I ran into the same problem playing around with 10,
>local file worked, loghost didn't and that was the problem.

>I don't remember earlier versions automatically adding loghost to /etc/hosts
>defining the machine itself, but it's been a while. 10 seems to do it
>automatically.

Good suggestion!  My Jumpstart post-install script removes `loghost'
from /etc/inet/hosts, so that shouldn't have been a problem.  However,
I found another one in /etc/inet/ipnodes.  That was the culprit.  It's
removed the same way now.

Disabling svc:/system/system-log:default and running syslogd with the
`-d' option revealed the error:

	# /usr/sbin/syslogd -d
	main(1): Started at time Thu Feb  3 19:13:35 2005
	hnc_init(1): hostname cache configured 128 entry ttl:600
	getnets(1): found 1 addresses, they are: 0.0.0.0.2.2
	amiloghost(1): testing 130.179.16.33.2.2
	conf_init(1): I am loghost

-- 
-Gary Mills-    -Unix Support-    -U of M Academic Computing and Networking-