As part of HCX Connector or Cloud Manager upgrade to version 4.3.0, the database will be transitioned from MongoDB to PostgreSQL. The process will ensure the conversion is complete and data is validated.
Thorough verification of the configuration and services in the HCX deployment is strongly recommended immediately after the upgrade is completed.
If there is any problem during the process, the upgrade will be canceled and the previous version running will be left active in the system. There will be no indication in the HCX UI regarding the failed upgrade and log analysis may be required.
VMware HCX
As a best practice, a procedure is made available in the unlikely event that any data inconsistency issues are found after the upgrade to HCX 4.3.0 is completed. Reverting the upgrade should be considered only as a last resource, after a potential issue is fully validated and for which no workaround or other remediation process has been identified.
# mount -t ext3 /dev/sda1 /recovery # findmnt --mountpoint /
TARGET SOURCE FSTYPE / /dev/sda3 ext3
TARGET SOURCE FSTYPE / /dev/sda2 ext3
# mount -t ext3 /dev/sda2 /slot1
# mount -t ext3 /dev/sda3 /slot2
# sh /opt/vmware/bin/recover.sh
The system will reboot at the end of the restoration process.
Once the Connector or Cloud Manager is back online, verify the version running, all configuration and services.
Refer to KB74869 (section 2 only) for instructions on how to remove the 4.3.0 plug-ins from vCenter and restart app-engine from the HCX Administration UI to reinstall the previous version plug-ins.
IMPORTANT NOTES: