Index: [Article Count Order] [Thread]

Date:  Wed, 11 Apr 2007 15:58:15 -0400
From:  "Paul Aviles" <paul.aviles (at mark) nickelnetworks.com>
Subject:  [coba-e:09530] Re: Recent server updates
To:  <coba-e (at mark) bluequartz.org>
Message-Id:  <200704112042.l3BKfpAG008750 (at mark) srv1.nickelnetworks.com>
In-Reply-To:  <038b01c77c52$c86911c0$3446460a@CIIC2>
X-Mail-Count: 09530

Michael,  I have seen some weird cced issues lately but not sure what is
causing this. But, my servers have been acting strange for a while. I tried
this and did work manually starting services in this order (network, ssh and
then cced) without any apparent problems.
 
You could change the runlevel values in /etc/rc.d/rc3.d for the startup
scripts. Here are the ones I have, just make sure you test this first on a
qa server. 
 
S00bluequartz
S00microcode_ctl
S06cpuspeed
S06dbrecover
S08iptables
S10-fix-raid-rebuild
S10network
S12syslog
S13irqbalance
S13named
S15cced.init
S15mdmonitor
S25netfs
S26lm_sensors
S28autofs
S40smartd
S55sshd
S56rawdevices
S56xinetd
S58ntpd
S60lpd
S64mysqld
S65dhcpd
S65dovecot
S74admserv
S80sendmail
S85gpm
S85httpd
S90crond
S95lcdsleep.init
S97messagebus
S98haldaemon
S99local
S99netvault

Paul

 
  _____  

From: Michael Aronoff [mailto:ma (at mark) ciic.com] 
Sent: Wednesday, April 11, 2007 12:02 PM
To: coba-e (at mark) bluequartz.org
Subject: [coba-e:09526] Re: Recent server updates


Paul wrote:
> There are apparently new server updates, or at least one of our client
servers picked some
> of them recently and got stuck doing several cced.init restarts
 
I did not notice this but has anyone else had cced.init take a VERY long
time on a reboot since this patch? 
 
I rebooted a server yesterday and it did not appear to come back up.  I
actually drove out to my datacenter so I could watch it boot.  It was
hanging on cced.init.  I wound up renaming /etc/init.d/cced.init to
cced.init2 and getting the server up.  If fully booted (with errors because
cced.init was not running), I then started cced.init manually and restarted
httpd so websites were working at least.  Then while looking at setting up a
new server cced.init finished and I restarted sendmail and all was ok.
 
I left it that way for the day and then at 10pm I restarted the machine
again and sure enough it took almost 30 min to boot, hanging at cced.init
for about 25 min.  Then it came back up fine.  So has anyone else seen this?
 
Also on a different note how can I move SSH and webmin in the startup
sequence so that if this happens again I can at least get into the server
from remote and see what is going on.  It seems to me that cced.init should
start later in the process, just before the web, admin and mail processes
that need it but after ssh, webmin, named and other processes that do not.
 
Thanks,
____________________________________
Michael Aronoff - CIIC, Inc. - Calabasas, CA


	

9530_2.html (attatchment)(tag is disabled)