Mailing List Archive: conserver-7.2.4 host rest

I am currently running conserver-7.2.4 on a Redhat 7.3 system. I was
wondering if you've heard of this issue I've ran into and if you know of
any remedy.

I console to a server attached to the console server and I am
installing the OS (solaris 9) through the console. After the the install
is finished, I go ahead and install various patches through a
patch-script. Before the patch install is finished, the console notifies
me that the host reset and it looks like it is rebooting the system. Is
there a buffer problem or some sort with this version? Or maybe the
output is coming on to the console screen so I can't see that the patch
has finished? Here is the output of the log for conserver:

+ ls -lL /a//usr/t/114014-01 ^M
total 138^M
-rwxr-xr-x 1 root root 262 Nov 7 2002 patchinfo^M
-rw-r--r-- 1 rroeotb o o rtoiot n g . . . ^M65
361Resett
^MLOM event: +1d+1h4m0s host reset
^Ming ... ^M
^M
p[-- MARK -- Tue May 6 16:58:00 2003]^M
^M
Netra T1 200 (UltraSPARC-IIe 500MHz), No Keyboard^M
OpenBoot 4.0, 256 MB memory installed, Serial #12633967.^M
Ethernet address 8:0:20:fc:ba:1f, Host ID: 80fcba1f.^M
^M

Thank you for your help.

Judy

On Thu, May 08, 2003 at 03:46:58PM -0700, Judy Lin wrote:
> I console to a server attached to the console server and I am
> installing the OS (solaris 9) through the console. After the the install
> is finished, I go ahead and install various patches through a
> patch-script. Before the patch install is finished, the console notifies
> me that the host reset and it looks like it is rebooting the system. Is
> there a buffer problem or some sort with this version? Or maybe the
> output is coming on to the console screen so I can't see that the patch
> has finished? Here is the output of the log for conserver:

there isn't a buffer problem or anything that i know of that would
cause what you are seeing. does your patch-script issue a reboot at
the end? i suppose it could be that the script finishes and goes about
rebooting before the output is able to come across the 9600 baud
connection. seems unlikely to me, but it's the only thing that i can
think of right now that would fit the output you sent. if it is the
case that it issues a reboot, you could just put something like a
'sleep 10' right before it and see if that fixes things up for you. if
so, it has something to do with the serial connection being so slow
(the 'rebooting...' message that got mixed with the 'ls' output is
probably coming over stderr instead of stdout, so it's able to sneak
out before the other and cause the mixed up output - which also leads
me to think you've got a reboot happening before all the ls output is
flushed).

anyway, i hope that helps in some way.

Bryan