CSPC upgrade from 2.7.3 to 2.7.4 fail ?

[Mark Wentz]

My main collector has been trying to perform a scheduled upgrade from 2.7.3 to 2.7.4 for some time and would fail due to the need for software download authorization.  Today i manually set off the upgrade and was requested to accept the download agreement.  i did and the system started downloading software and eventually upgrading.  

I stated getting messages that it could not get updated information...  I don't know if the upgrade completed or failed.  I can get into the system but when i check anything on the software page i get "Unable to get data from server for report: ..........."

[Jarrett Pomeroy]

Hello Mark,



The install process may still be running if you recently started the upgrade process. From the "admin" CLI can you run the command "show apply" and see if it's showing as completed now? If it does show that the upgrade has completed, please try restarting the adminshell by running the command "service adminshell restart" from the root username and try reloading the Software Update Tab.



Thank you,

Jarrett

[Mark Wentz]

Jarrett,

the GUI is finally reporting the status and it agrees with my findings from the CLI:   Status:  Apply-in-progress...

It's been in this status since Tuesday....   Where do i go from here ?

Thanks for your help.

[Jarrett Pomeroy]

Hello Mark,

Can you please check and see if the re-applying steps found in the 2.7.3.2 release notes helps to resolve your issues? The steps start on page 6 in “known issue”.

https://www.cisco.com/c/dam/en/us/support/docs/cloud-systems-management/common-services-platform-collector-cspc/CSPC-ReleaseNotes-2732.pdf

Thank you,

Jarrett

[Mark Wentz]

Jerrett,

In checking my system before i attempted the Bug Fix i noted that the system status is now reporting "Downloaded" with no other options. 

Here is what i loaded for the bug fix and what was returned:

[root@Collector1 ~]# /bin/sh/opt/LCM/bin/sqliteToDerbyScripts_shell.sh "connect 'jdbc:derby://localhost:1527/opt/LCM/info/updates; create=true;user=app;password=password';" "update updates set PACKAGE_STATUS = 'Apply-failed' where PACKAGE_STATUS ='Apply-in-progress';"

-bash: /bin/sh/opt/LCM/bin/sqliteToDerbyScripts_shell.sh: Not a directory

[root@Collector1 ~]#

Thanks,

Mark

[CC Cisco]

You're missing a space in your command:

"/bin/sh/opt/LCM"

[CC Cisco]

Correct command (one line only):

/bin/sh /opt/LCM/bin/sqliteToDerbyScripts_shell.sh "connect 'jdbc:derby://localhost:1527/opt/LCM/info/updates; create=true;user=app;password=password';" "update updates set PACKAGE_STATUS = 'Apply-failed' where PACKAGE_STATUS = 'Apply-in-progress';"

[Mark Wentz]

Thanks for the reply and correction.  My system accepted the command but did nothing as the status had already changed to Download state.

I then attempted to follow the manual upgrade information in the guide.  I'm a bit confused as the instructions are to apply them from root but I could only get them to work from the Admin signon.  

My system is once again "applying" the upgrade.  We'll see what it looks like tomorrow.

Thanks for your continued support.

[snarla]

Hi mark,

Can you please provide an update on if you were able to upgrade to 2.7.4 or still having issues?

Best Regards,

Sateesh

[markusscansource]

Sorry, i have been OOP.   Following instructions i attempted to manually apply the patches.  This is what i get on the first one: 

[root@Collector1 ~]# tail -f /opt/LCM/logs/apply
[Thu May  3 14:01:31 EDT 2018] status db updation completed
status before trim : Applied
status after trim : Applied
inside
[Thu May 17 13:32:58 EDT 2018] Apply TCL Called with arguments "JeOS" and "/opt/                                             LCM/tmp/jeos-30.0.4-0-lnx64.zip"
[Thu May 17 13:33:55 EDT 2018] Executing pre-install scripts
[Thu May 17 13:33:58 EDT 2018] Executed pre-install scripts
[Thu May 17 13:33:58 EDT 2018] Executing install script for "Hardened CentOS"
[Thu May 17 13:35:45 EDT 2018] Executed install script for "Hardened CentOS"
[Thu May 17 13:35:45 EDT 2018] Install script for "Hardened CentOS" returned failure

[Justin Sprake]

Hello,

Could you provide the command which you executed to initiate the install of the jeos-30.0.4-0-lnx64.zip file?

[markusscansource]

nohup /opt/LCM/bin/lcmagent-apply.tcl JeOS /opt/LCM/tmp/jeos-30.0.4-0-lnx64.zip

[Justin Sprake]

Hi Mark,

Could you re run the command again but be sure to include the '&' symbol at the end? This will ensure the command is executed in the background. Additionally, could you confirm the size of the file as well via '#ls -larths /opt/LCM/tmp/jeos-30.0.4-0-lnx64.zip'

[snarla]

Hi Justin,

Please use below command as it will push the installation to background so it will not stop upon session timeout.

/opt/LCM/bin/lcmagent-apply.tcl JeOS jeos-30.0.4-0-lnx64.zip &

[Mark Wentz]

I processed the JeOS and the SP for 2.7.4 . . . from root

My system software update screen shows current version 2.7.4 and that 2.7.4.1 was downloaded but the Upgrade Now button was grayed out.

Installed updates shows last version installed Succssfully is 2.7.3 . . .

manually (from Admin) started the upgrade to 2.7.4.1 and will report back tomorrow.  Currently the Status is Apply-in-progress  since this morning.  I have a feeling it is once again stuck like 2.7.4