Is an LMS 3.2 reinstall in order?

Answered Question
Jan 11th, 2010

Upgrading another LMS 3.1 instance to 3.2 resulted in the following. Do the ERRORs warranty blowing LMS away and reinstall 3.2 from scratch?

===================- Possible Warnings/Errors Encountered -===================
WARNING: at group permissions needed for casuser.
WARNING: Set at for casuser in /etc/cron.d/at.allow Or
WARNING: Unset at for casuser in /etc/cron.d/at.deny.
WARNING: cron group permissions needed for casuser.
WARNING: Set cron for casuser in /etc/cron.d/cron.allow Or
WARNING: Unset cron for casuser in /etc/cron.d/cron.deny.
WARNING: Ensure that you have installed the recommended Solaris 10 cluster patches released
WARNING: on  Apr/17/07, in this server.
WARNING: If these cluster patches are not installed, please download and install them
WARNING: from http://www.sun.com/.
WARNING: Otherwise, some features of the CiscoWorks applications will not function properly.
WARNING: Exiting installation beyond this point might result in system instability.
ERROR: /etc/rc.config.d
ERROR:    group name <50053> not found in group table(s)
ERROR:    owner name <2454> not found in passwd table(s)
WARNING: /opt/CSCOpx/objects/perl/bin/perl5.8.7 <no longer a linked file>
WARNING: /opt/CSCOpx/objects/perl/bin/pstruct <no longer a linked file>
WARNING: /opt/CSCOpx/objects/perl/bin/s2p <no longer a linked file>
WARNING: An error occurred while starting the TFTP server.
chmod: WARNING: can't access /opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/cmf/DeviceData.zip
ERROR: attribute verification of </opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/cmf/DeviceData.zip> failed
WARNING: RME device package install has failed.
WARNING: Please copy the device packages manually from /opt/vgi/ehealth/backups/lms32/lms32install/install/cids_pkgs to /opt/CSCOpx/MDC/tomcat/webapps/rme/WEB-INF/lib/pkgs
WARNING: RME device package install has failed.
WARNING: Please copy the device packages manually from /opt/vgi/ehealth/backups/lms32/lms32install/install/cids_pkgs to /opt/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/rme
   ERROR: /usr/sbin/in.tftpd is not installed.
WARNING: '/etc/inetd.conf' already contains an entry for tftp server.
ERROR: No tftp server detected, yet it is configured in '/etc/inetd.conf'.
ERROR: Software Image Manager requires a tftp server.
ERROR: Contact your local technical support representative to install
ERROR: a tftp server on this machine.
WARNING: An error occurred while starting the tftp server.
WARNING: Contact your local technical support representative to start the tftp server.
WARNING: To ensure that you retain the latest device support,
WARNING: please install the latest Device Packages from Cisco.com @
WARNING: http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-rme
WARNING: Please refer to the Installation and Setup Guide for details.
WARNING: To ensure that you retain the latest device support and bug fixes, please install
WARNING: the latest Service Pack (SP) for Campus Manager 5.2
WARNING: You can download the latest SP from Cisco.com, at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-campus.
WARNING: Please refer to the Installation and Setup Guide for details.
WARNING: User Tracking Utility 1.0 is not compatible with Campus Manager 5.2
WARNING: User Tracking Utility 1.1.1 is available for Campus Manager 5.2
WARNING: You can download the latest UTU from Cisco.com, at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-campus-crypto.

Correct Answer by Joe Clarke about 7 years 1 month ago

You absolutely have a bad disc or drive:

cp: /opt/vgi/ehealth/backups/lms32/lms32install/install/DeviceData/DeviceData.zip: read: No such device or address

This is a sure sign of corruption.

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (2 ratings)
Loading.
yjdabear Mon, 01/11/2010 - 11:21

ls -al /opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/cmf/DeviceData.zip

/opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/cmf/DeviceData.zip: No such file or directory

less /etc/inetd.conf

tftp    dgram   udp     wait    root    /opt/CSCOpx/bin/in.tftpd        in.tftpd -s /tftpboot
tftp   dgram   udp6    wait    root    /usr/sbin/in.tftpd              in.tftpd -s /tftpboot

Joe Clarke Mon, 01/11/2010 - 11:45

Yes, the one error which is really bad is the error about RME package installation.  Simply copying the packages will not be sufficient.  Given this, I would do a 3.2 install from scratch, then restore your 3.1 backup.

yjdabear Mon, 01/11/2010 - 12:20

Thanks. Should I clean up the following residue before proceeding with the LMS 3.2 install?

/opt/CSCOpx)->ls -alR
.:
total 16
drwxrwxr-x   5 casuser  casusers    1024 Jan 11 15:14 .
drwxr-xr-x   7 casuser  casusers    1024 Dec 19  2008 ..
-rw-r--r--   1 root     other       1806 Jan 11 10:50 .odbc.ini.orig
drwx------   2 casuser  casusers    1024 Feb  3  2009 .ssh
drwxr-x---   3 casuser  casusers    3072 Jan 11 15:14 bin
drwxr-xr-x   2 root     other         96 Oct 23  2008 log

./bin:
total 10
drwxr-x---   3 casuser  casusers    3072 Jan 11 15:14 .
drwxrwxr-x   5 casuser  casusers    1024 Jan 11 15:14 ..
drwxrwxr-x   4 casuser  casusers    1024 Jan 11 15:13 RestoreTools

./bin/RestoreTools:
total 12
drwxrwxr-x   4 casuser  casusers    1024 Jan 11 15:13 .
drwxr-x---   3 casuser  casusers    3072 Jan 11 15:14 ..
drwxr-x---   2 casuser  casusers    1024 Jan 11 14:51 ipm
drwxr-x---   2 casuser  casusers    1024 Jan 11 15:11 rmeng

./bin/RestoreTools/ipm:
total 110
drwxr-x---   2 casuser  casusers    1024 Jan 11 14:51 .
drwxrwxr-x   4 casuser  casusers    1024 Jan 11 15:13 ..
-rwxr-x---   1 casuser  casusers   53578 Jan  6  2009 ipmRestoreLib.pm.default

./bin/RestoreTools/rmeng:
total 196
drwxr-x---   2 casuser  casusers    1024 Jan 11 15:11 .
drwxrwxr-x   4 casuser  casusers    1024 Jan 11 15:13 ..
-rw-r--r--   1 root     other      19618 Nov 20  2008 ccrreg.sh
-rwxr-x---   1 casuser  casusers   77474 Jan 24  2008 rmengRestoreLib.pm.default

./log:
total 8
drwxr-xr-x   2 root     other         96 Oct 23  2008 .
drwxrwxr-x   5 casuser  casusers    1024 Jan 11 15:14 ..
-rw-r--r--   1 root     other       2922 Jan 11 10:55 schema.log

Joe Clarke Mon, 01/11/2010 - 12:29

I typically destroy NMSROOT before doing a clean installation.  You're using symlinks, but the same principle applies.  Remove the /opt/CSCOpx symlink, and destroy the physical NMSROOT directory.

yjdabear Mon, 01/11/2010 - 13:37

Failed again, despite a clean install:

===================- Possible Warnings/Errors Encountered -===================
WARNING: at group permissions needed for casuser.
WARNING: Set at for casuser in /etc/cron.d/at.allow Or
WARNING: Unset at for casuser in /etc/cron.d/at.deny.
WARNING: cron group permissions needed for casuser.
WARNING: Set cron for casuser in /etc/cron.d/cron.allow Or
WARNING: Unset cron for casuser in /etc/cron.d/cron.deny.
WARNING: Ensure that you have installed the recommended Solaris 10 cluster patches released
WARNING: on  Apr/17/07, in this server.
WARNING: If these cluster patches are not installed, please download and install them
WARNING: from http://www.sun.com/.
WARNING: Otherwise, some features of the CiscoWorks applications will not function properly.
WARNING: Passwords do not match. Please try again.
WARNING: Exiting installation beyond this point might result in system instability.
ERROR: Cannot resolve symbolic link '/tftpboot' for use by the tftp server.
ERROR: Create the directory linked to by '/tftpboot'.  It should have permissions of 'rwxrwxrwx'.
WARNING: An error occurred while starting the TFTP server.
chmod: WARNING: can't access /opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/cmf/DeviceData.zip
ERROR: attribute verification of failed
WARNING: RME device package install has failed.
WARNING: Please copy the device packages manually from /opt/vgi/ehealth/backups/lms32/lms32install/install/cids_pkgs to /opt/CSCOpx/MDC/tomcat/webapps/rme/WEB-INF/lib/pkgs
WARNING: RME device package install has failed.
WARNING: Please copy the device packages manually from /opt/vgi/ehealth/backups/lms32/lms32install/install/cids_pkgs to /opt/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/rme
        ERROR: /usr/sbin/in.tftpd is not installed.
WARNING: '/etc/inetd.conf' already contains an entry for tftp server.
ERROR: No tftp server detected, yet it is configured in '/etc/inetd.conf'.
ERROR: Software Image Manager requires a tftp server.
ERROR: Contact your local technical support representative to install
ERROR: a tftp server on this machine.
ERROR: Cannot resolve symbolic link '/tftpboot' for use by the tftp server.
ERROR: Contact your local technical support representative to create the directory
ERROR: linked to by '/tftpboot'.  It should have permissions of 'rwxrwxrwx'.
WARNING: An error occurred while starting the tftp server.
WARNING: Contact your local technical support representative to start the tftp server.
WARNING: To ensure that you retain the latest device support,
WARNING: please install the latest Device Packages from Cisco.com @
WARNING: http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-rme
WARNING: Please refer to the Installation and Setup Guide for details.
WARNING: To ensure that you retain the latest device support and bug fixes, please install
WARNING: the latest Service Pack (SP) for Campus Manager 5.2
WARNING: You can download the latest SP from Cisco.com, at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-campus.
WARNING: Please refer to the Installation and Setup Guide for details.
WARNING: User Tracking Utility 1.0 is not compatible with Campus Manager 5.2
WARNING: User Tracking Utility 1.1.1 is available for Campus Manager 5.2
WARNING: You can download the latest UTU from Cisco.com, at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-campus-crypto.
=====================================================================

from /var/tmp/Ciscoworks_install__154100.log

log4j:WARN No appenders could be found for logger (com.liferay.portlet.devsupport.job
registration.SystemPSUJobCreationImpl).
log4j:WARN Please initialize the log4j system properly.
Unable to read next job id from JRM table... writing sentinel query
writeSentinel() - max job id is 0
writeSentinel() - job id has been made 1000.
writeSentinel() - wrote sentinel with id 1001
Job id to be used next is 1001
Registering Device_Support with job id 1001...
JobRegistrationUtil : Completed successfully.
    Stopping database engine cmfEng
SQL Anywhere Command File Hiding Utility Version 10.0.1.3830
INFO: After create System jobs
INFO: Installing /opt/vgi/ehealth/backups/lms32/lms32install/install/DeviceData/Devic
eData.zip
cp: /opt/vgi/ehealth/backups/lms32/lms32install/install/DeviceData/DeviceData.zip: re
ad: No such device or address
INFO: Could not copy the DeviceData.zip file
chown: /opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/pkgs/cmf/DeviceData.zi
p: No such file or directory
chmod: WARNING: can't access /opt/vgi/csco/CSCOpx/www/classpath/com/cisco/nm/xms/psu/
pkgs/cmf/DeviceData.zip
INFO: cwportal.log already exists
INFO: Updating the permission for cwportal.log file
INFO: Going to set the Default Community
INFO: executing /opt/CSCOpx/bin/cwjava -cw /opt/CSCOpx -cp:p MDC/tomcat/webapps/cwpor
tal/WEB-INF/classes:MDC/tomcat/webapps/cwportal/WEB-INF/lib/ext-impl.jar:MDC/tomcat/w
ebapps/cwportal/WEB-INF/lib/log4j.jar com.liferay.portal.util.ciscoWorks.SetDefaultCo
mmunity LMS
INFO:after executing /opt/CSCOpx/bin/cwjava -cw /opt/CSCOpx -cp:p MDC/tomcat/webapps/
cwportal/WEB-INF/classes:MDC/tomcat/webapps/cwportal/WEB-INF/lib/ext-impl.jar:MDC/tom
cat/webapps/cwportal/WEB-INF/lib/log4j.jar com.liferay.portal.util.ciscoWorks.SetDefa
ultCommunity LMS
INFO:Before calling installf for devicedata.zip
ERROR: attribute verification of /psu/pkgs/cmf/DeviceData.zip> failed
    pathname does not exist
INFO:after calling installf for devicedata.zip
INFO : END -> postinstall of cwportal pkg...

Installation of was successful.

INFO: Package CSCOcwportal - Installed. Correcting pkgchk errors - Done.
application CSCOcwportal                 Portal engine
INFO: performing pkgadd -a /tmp/cscotmp/CSCO.admin -d /opt/vgi/ehealth/backups/lms32/
lms32install/install/packages CSCOcwa

Processing package instance from ll/install/packages>

CiscoWorks Assistant(sparc) 1.2
Using as the package base directory.
## Processing package information.
## Processing system information.
   16 package pathnames are already properly installed.
## Verifying disk space requirements.

Installing CiscoWorks Assistant as

I'm wondering if I have a corrupted ISO image on this box.

Correct Answer
Joe Clarke Mon, 01/11/2010 - 14:41

You absolutely have a bad disc or drive:

cp: /opt/vgi/ehealth/backups/lms32/lms32install/install/DeviceData/DeviceData.zip: read: No such device or address

This is a sure sign of corruption.

yjdabear Tue, 01/12/2010 - 08:22

Yes, indeed. Apparently the disc image that I've used *successfully*during the first upgrade has somehow gotten corrupted sitting on the SAN disk (that's rather disconcerting). The MD5 checksums are the same across all the copies, yet I can no longer access DeviceData.zip on any of them. After making a new image from the physical media again, it's finally good to go.

Actions

This Discussion