This is a known issue affecting VMware Cloud Foundation 2.3.x. There is currently no resolution.
Workaround:
Once you see the message upgrade failed at stage JRE_UPGRADE_STAGE in /home/vrack/lcm/logs/lcm.log on the SDDC Manager Controller VM, you can use the follwing steps to workaround the issue:
- Log in to the SDDC Manager Controller VM as the root user.
- Issue the following command to replace the cacerts file with a backup copy created during the upgrade:
cp /usr/java/jre-vmware/lib/security/cacerts.bak /usr/java/jre-vmware/lib/security/cacerts
- Remount the NFS mount from the SDDC Manager Utility VM by issuing the following command:
mount -o remount,nosuid,nodev,noexec /home/vrack/VMware/vRack/nfs-mount/lcm/repo
- Open the /etc/fstab file with a text editor.
vi /etc/fstab
- Find the line that reads
/home/vrack/VMware/vRack/nfs-mount/lcm/repo nfs defaults
and change defaults to nosuid,nodev,noexec.
Note: This change enforces more restrictive NFS mount permissions to make it more secure. Normally, this is done automatically in the third-party upgrade script.
- Save and close the file by typing :wq.
- Open the /home/vrack/lcm/lcm-app/conf/application-evo.properties file with a text editor:
vi /home/vrack/lcm/lcm-app/conf/application-evo.properties
- Find the line that reads
lcm.upgrade.thirdparty.skip=false
and change false to true.
- Save and close the file by typing :wq.
- Issue the following commands to restart the lcm service:
systemctl stop scs
systemctl stop lcm
systemctl start lcm
systemctl start scs
Note: The upgrade can be re-attempted at this point.