-
Título
Replication to another Core failing with error: "Object reference not set to an instance of an object" -
Descripción
There are 2 protected systems: Core running 6.1.3 is trying to replicate to another core running 6.2.0.17839, and error is generated "Object reference not set to an instance of an object.". Trying to point the replication to another Core, deletion and re-creation of the replication link is not resolving the issue. -
Causa
Source Core was upgraded to 6.2 but then downgraded from 6.2 to 6.1.3 version, leaving all registry values intact. Possible reason for downgrade: obsolete agents that stopped to work after Core has been upgraded. -
Resolución
First of all, the reason for such error is that when Core is upgrading to the new version, many registry values are changed and even removed as obsolete. Downgrade process is not really possible, unless all registry information and Core folders are wiped completely after new version is uninstalled in preparation for downgrade process.
If downgrade process is absolutely necessary, then please make sure that information is recorded before uninstall:
- Core license key file in .LIC format (retrieved from rapidrecovery.licenseportal.com)
- all repositories' paths, including all extensions (if more than 1)
- (optional) global retention policy and per-agent (if there are customized retention policies for some agents)
- (optional) agents' protection schedules
- export encryption keys to files (if encryption is used). Note: it's not possible to recreate the same encryption key even if passphrase is known! Encryption keys are having unique IDs that are tied to encrypted Recovery Points.The process in nutshell look like this:
- uninstall Rapid Recovery. Reboot if needed
- export to file (for backup purposes) and delete registry key HKLM\Software\Apprecovery
- delete "C:\Program Files\Apprecovery" and "C:\ProgramData\Apprecovery" folders- install previous version of the Rapid Recovery and apply license LIC file
- open existing DVM repositories one by one and wait for them to mount. Agents' list will be populated in the Recovery Points Only section
- re-protect them one by one using previously saved protection schedules or just use a new schedules that are suitable for server's nature (backups data when content is being created/changed on the server)
- change retention policy to what was previously or just keep default if it's suitable
- import encryption key files (if were used)Setup replication link and force replication to make sure that it's working now.
Note: it's highly recommended to open a Support ticket, so the process can be overlooked by Quest Support engineer and issues are addressed in timely fashion, if any