FMCv Upgrade Failed From 6.2.3 to 6.4.0

[-Sparrow-]

Hi there folks! 

I've run into an issue with my FMC upgrade.  Trying to go from 6.2.3 to 6.4.0 and It dies at 22%.  I originally started the upgrade from the GUI but since it failed I only have CLI access now.  I ran the upgrade_resume.sh as instructed from the prompt on the CLI and it came back with an error about the upgrade package not being in the /var/sf/updates.   Strange... Shouldn't it have been there when I originally started the upgrade from the GUI?  Anyway I re-downloaded the upgrade package from the Cisco support site and put it back in /var/sf/updates with WinSCP.

I ran the upgrade_resume.sh command and now its trying but dies at 22%. heres the output from the failure:

[200702 17:28:41:679] FAILED 300_os/075_extract_preinstall.sh
[200702 17:28:41:680] ====================================
[200702 17:28:41:681] tail -n 10 /var/log/sf/Cisco_Firepower_Mgmt_Center_Upgrade-6.4.0/300_os/075_extract_prein stall.sh.log

./functions.util: line 569: popd: directory stack empty
./functions.util: line 571: pushd: /new-root//usr/local/sf/htdocs/gwt-utils: No such file or directory
Deleting old timestamp dirs from gwt-utils
rm: refusing to remove '.' or '..' directory: skipping '.'
./functions.util: line 574: popd: directory stack empty
Found exactly one preinstall archive, continuing to extract it
untaring into /new-root
tar: -C: Cannot open: No such file or directory
tar: Error is not recoverable: exiting now
Fatal error: Couldn't untar the preinstall archive

[200702 17:28:42:276] MAIN_UPGRADE_SCRIPT_END
[200702 17:28:42:277] Fatal error: Error running script 300_os/075_extract_preinstall.sh
[200702 17:28:42:280] Exiting.
removed '/tmp/upgrade.lock/main_upgrade_script.log'
removed '/tmp/upgrade.lock/PID'
removed '/tmp/upgrade.lock/LSM'
removed '/tmp/upgrade.lock/status_log'
removed directory: '/tmp/upgrade.lock'
[200702 17:28:42:282] Attempting to remove upgrade lock
[200702 17:28:42:283] Success, removed upgrade lock
Process 29181 exited.I am going away.
RC: 256
The update failed!
Removed MAKESELF script: /var/tmp/upgrade-patch/Cisco_Firepower_Mgmt_Center_Upgrade-6.4.0-113.sh
root@firepower:/#

Does anyone have any thoughts before I bite the bullet and do as TAC says which is to reimage?  I have a good backup on 6.2.3 so if i have to reimage it wouldnt be that bad but i'd like to be able to salvage what I have if possible.

Thanks!

[Marvin Rhoads]

If TAC has already had a look at it and advises to reimage, then I would follow their advice. They have access to more advanced analysis and troubleshooting tools than anybody who's helping out here.

[-Sparrow-]

Hi Marvin,

That is partly the problem.  TAC hasn't taken a look at it live.  The assigned engineer has looked at some log files that he requested that I send to him.  He said the DB was corrupted and I need to reimage.  That's about it.  I figured I'd ask here if anyone had any ideas. 

The engineer sent a link to the reimage document that doesn't work with our vFMC.  I took a look and I don't have the option in LILO to boot from System_Restore.  I also didn't have the same BIOS options that needed to be set in the Troubleshooting section of the reimage document.

The FMC is running, I just can't get access in the GUI.  I do have CLI access.

I'm stuck in the first part of the reimage process.  Should I just spin up a new one from the .ova, restore settings from my good backup and re-host the license?

Thanks!

[-Sparrow- · Accepted solution]

The reimage wasn't going well for me...TAC provided incorrect info for the reimage.

Ended up Spinning up a new vFMC and rehosting licenses, reregistering the SFR module with the new vFMC, and restoring vFMC backup.  

Everything is back to square one now v6.2.3.  I'll try the FMC upgrade to 6.4.0 again in the near future.