From: John L. Lanier (jllanier@parsec.com)
Date: Mon Jan 31 2005 - 03:20:23 EST
Hello,
Thanks to replies from the following people:
Edward [ed1901@hotmail.com]
Bill Hassell [bill@billhassell.com]
Mohan Sundaram [Mohan_Sundaram@emitac.ae]
Bob Vance [bobvance@alumni.caltech.edu]
I have set my DISPLAY variable to "hostname:0.0" and have "hostname" (not the FQDN) defined in the /etc/rc.config.d/netconf file. After rebooting, I see no more weirdness with hanging commands and the sw-related commands failing.
Since this zx6000 did not come installed with a Management Processor (I didn't realize just how important it was to have this hardware until recently!), I will have to wait until I am in the office next to see if CDE is working as it should. I will also be looking into purchasing an MP in the near future.
Thanks again and see ya next time!
--John Lanier
-----Original Message-----
From: hpux-admin-owner@DutchWorks.nl [mailto:hpux-admin-owner@DutchWorks.nl <mailto:hpux-admin-owner@DutchWorks.nl> ]
On Behalf Of John L. Lanier
Sent: Friday, January 28, 2005 7:22 PM
To: hpux-admin@DutchWorks.nl
Subject: [HPADM] Strange behavior seen after a cold install of 11i V2 on a
zx6000 workstation
Hello,
I'm somewhat new to HPUX (I work primarily in Tru64 and Linux land), so bear
with me.
Config.
=====
System: series model ia64 hp workstation zx6000 system.
OS version: HP-UX version B.11.23 (11i V2)
Keyboard and mouse are both USB
Monitor: Sony 21 in. Trinitron GDM2038
We just received shipment of this system about two weeks ago, minus the
monitor.
Problem:
======
I recently did a cold install of 11i V2 on this system and all
appeared well afterwards, until I tried to do a "swverify" and I get this
error:
--> swverify
======= 01/27/05 13:29:12 MST BEGIN swverify SESSION
(non-interactive) (jobid=merc72.parsec.com-0013)
* Session started for user "root@merc72.parsec.com".
* Beginning Selection
ERROR: Could not contact host "merc72.parsec.com". Make sure the
hostname is correct and an absolute pathname is specified
(beginning with "/").
* Target connection failed for "merc72.parsec.com:/".
ERROR: More information may be found in the daemon logfile on this
target (default location is
merc72.parsec.com:/var/adm/sw/swagentd.log).
* Selection had errors.
======= 01/27/05 13:29:12 MST END swverify SESSION (non-interactive)
(jobid=merc72.parsec.com-0013)
*******
I also get the same types of errors with "swlist" and "swinstall".
Looking at the /var/adm/sw/swagentd.log, I see the following:
WARNING: An attempt to get the network host entry for
"merc72.parsec.com" failed. This may result in denial of
access to users and agents at this host. Check the spelling
of this name, then your "/etc/hosts" file, or your
"/etc/resolv.conf" file and DNS resolver configuration. The
nslookup program may be helpful in isolating this problem.
Well, my hosts file looks OK:
====================
#
127.0.0.1 localhost loopback
10.100.0.61 merc72.parsec.com merc72
10.100.0.25 email.parsec.com email
So does /etc/resolv.conf:
=================
domain parsec.com
nameserver 10.100.0.25
nslookup doesn't report any errors:
========================
.........
Trying DNS
Trying NIS
looking up FILES
Name: merc72.parsec.com
Address: 10.100.0.61
Aliases: merc72
*********
NOTE: traceroute/ping/telnet all work fine to/from this system.
Here's where it gets interesting:
=====================
I remembered seeing some network weirdness in the past when entries in the
/etc/rc.config.d/netconf are not right.
I checked my file and see this:
====================
.......
HOSTNAME="merc72.parsec.com"
OPERATING_SYSTEM=HP-UX
LOOPBACK_ADDRESS=127.0.0.1
.......
INTERFACE_NAME[0]="lan1"
IP_ADDRESS[0]="10.100.0.61"
SUBNET_MASK[0]="255.255.255.0"
BROADCAST_ADDRESS[0]=""
INTERFACE_STATE[0]=""
DHCP_ENABLE[0]=0
.......
ROUTE_DESTINATION[0]="default"
ROUTE_MASK[0]=""
ROUTE_GATEWAY[0]="10.100.0.1"
ROUTE_COUNT[0]="1"
ROUTE_ARGS[0]=""
.......
All of the IP addresses (excluding loopback) are correct for our site.
My routing tables look OK:
===================
--> netstat -rn
Routing tables
Destination Gateway Flags Refs Interface Pmtu
127.0.0.1 127.0.0.1 UH 0 lo0 4136
10.100.0.61 10.100.0.61 UH 0 lan1 4136
10.100.0.0 10.100.0.61 U 2 lan1 1500
127.0.0.0 127.0.0.1 U 0 lo0 0
default 10.100.0.1 UG 0 lan1 0
I checked another HPUX system and it has just the hostname, not the FQDN.
So, I shortened my entry to just the hostname, rebooted, and all of the
sw-related commands work as expected now.
BUT, as result of doing this, I am now noticing the following:
1. Commands appear to hang sporadically for a few seconds before completing
2. I can't login through the GUI/CDE; the display just loops between the HP
banner
and the gray screen right before CDE launches. I never actually see a
CDE banner.
When I set the hostname back to the FQDN in netconf and reboot, commands no
longer
hang and I actually get a CDE banner login. BUT, when I try to login to the
CDE I get an error saying that I am not able to launch CDE and it recommends
that I launch a failsafe session to troubleshoot. And, of course, the
sw-related commands fail again.
I also see the following in the $HOME/.dt/errorlog for root:
...........
*** INFORMATION(0): dtwm: PID 2546: Wed Jan 26 15:55:01 2005
Workspace Manager: I/O error on display:: :0.0
*** [109]
Questions:
==========
1. If I set the $DISPLAY variable to "hostname:0.0", will that solve the
CDE problem?
2. What about the weird command hang behavior when the hostname is
shortened in /etc/rc.config.d/netconf?
3. Why do the sw-related commands fail when the FQDN is defined in netconf?
Thanks in advance for any help/advice you can offer.
--John Lanier
--
---> Please post QUESTIONS and SUMMARIES only!! <---
To subscribe/unsubscribe to this list, contact majordomo@dutchworks.nl
Name: hpux-admin@dutchworks.nl Owner: owner-hpux-admin@dutchworks.nl
Archives: ftp.dutchworks.nl:/pub/digests/hpux-admin (FTP, browse only)
http://www.dutchworks.nl/htbin/hpsysadmin (Web, browse & search)
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 11:02:45 EDT