Problems after CVSUP

[Edward Carmody]

Conrad,

Run mergemaster a second time?  Didn't do that...

I've since rebooted this box a couple times...is it too late to run
mergemaster a second time?  Also, is it too late to run the make all install
in /etc/mail?

If it's too late for these, how do I fix the damage I've done?

-----Original Message-----
From: Conrad Sabatier [mailto:conrads at cox.net] 
Sent: Saturday, May 29, 2004 2:16 AM
To: Edward Carmody
Cc: freebsd-questions at freebsd.org
Subject: Re: Problems after CVSUP


On 29-May-2004 Edward Carmody wrote:
> Not quite a newbie, but not very good either, apparently. :(  I did a
cvsup
> last night, following instructions from the handbook, and hosed up my box.
> 
> =================================
> [ecarmody at oberon /home/ecarmody] $ uname -a
> FreeBSD oberon.hudsonvalleynetworks.com 5.2.1-RELEASE-p7 FreeBSD
> 5.2.1-RELEASE-p7 #0: Thu May 27 19:57:55 EDT 2004
> root@:/usr/obj/usr/src/sys/GENERIC  i386
> [ecarmody at oberon /home/ecarmody] $
> =================================
> 
> I should not be running 5.2.1-RELEASE due to my inexperience, I know, but
> need to in order to support my IDE controller.  I was using the box for
www,
> sendmail, tacacs, ntp, syslog, samba, for my home network, and just trying
> to learn UNIX a la FreeBSD.  
> 
> My immediate issues are these:
> At boot, ntpd hangs, holding up the rest of the startup process.  If I
kill
> it with a ctrl-c, sendmail then does some funky things, and also hangs.  
> 
> ===================
> /var/log/ntp shows:
> [root at oberon /var/log] # cat ntp.log 
> 28 May 23:53:37 ntpd[609]: ntpd exiting on signal 15
> 28 May 23:56:37 ntpd[409]: ntpd exiting on signal 2
> [root at oberon /var/log] #
> 
> which I think is from me killing it via ctrl-c
> ===================
> 
> So, to workaround temporarily, I commented out ntpd_enable="YES" in
rc.conf.
> Ntpd obviously doesn't hold up my boot anymore but sendmail does.  
> 
> ===================
> Starting sendmail.
> safefile(etc/mail/services.switch, uid=0, gid=0, flags=6480, mode=400):
> safedirpath(/etc/mail, uid=0, gid=0 flags =6580, level=0, offset=0):
>       [dir /etc/mail] OK
>       No such file or directory
> 
> Then I get similar lines for /etc/mail/submit.cf, etc/mail/relay-domains,
> The last line is:
> Recipient names must be specified
> 
> I can't seem to find the output I'm seeing in /var/log/messages or
> /var/log/maillog, so I've hand-typed in the bit above.
> 
> /var/log/maillog has lots of:
> May 29 00:00:05 oberon sm-msp-queue[436]: i4S2ug7P087593: to=root,
> ctladdr=root (0/0), delay=1+01:03:23, xdelay=00:00:00, mailer=relay,
> pri=4980184, relay=[127.0.0.1], dsn=4.0.0, stat=Deferred: Connection
refused
> by [127.0.0.1]
> 
> which also started after my cvsup last night.  
> ===================
> 
> 
> Not sure what else to provide.  Any direction on how to 'shoot and fix
this
> wins a free beer next time you're in NYC.  Thanks...

Did you forget to run mergemaster?

When updating via cvsup, you should *always* do the following:

1) cvsup
2) make buildworld
3) make buildkernel
4) make installkernel
5) reboot in single-user mode
6) run mergemaster -p
7) make installworld
8) run mergemaster

Then, double-check /etc/rc.conf against /etc/defaults/rc.conf to see if
there
are any important changes in startup functionality.  Then, and only then,
should it be safe to reboot.

If you've merged in any changes under /etc/mail, you should also cd
/etc/mail
and make all install before rebooting as well.

HTH

-- 
Conrad Sabatier <conrads at cox.net> - "In Unix veritas"