MID Server - Stuck Validating Post-Clone

[dannykamins]

Hi Everyone,

After troubleshooting, I found this information available via KB0746702, but hopefully presenting this in its own post may save other admins some trouble.

My team recently cloned over our Development instance and attempted to rely on the "Preserve Users and Related Tables" checkbox of the clone profile. We found that this still does not work in the Yokohama release, and I am unsure if there is a plan to include the functionality in future releases. For most of our user records this was not a concern, but the clone did impact the MID Server User(s) in our instance and clear their set passwords.

This resulted in authentication failures between our instance and our MID Servers, and stopping the service on the underlying VMs. New passwords were set from the instance, the config file found under C:\(MID Folder)\agent was updated to reflect the change, and one of our MIDs came back online. When the other stayed offline, we attempted to invalidate/revalidate and found that we were "stuck".

After trying (and failing) to start the MID Server service, I checked the agent logs to find the following message, and traced its source back to the new password that was set for the MID Server user:

"ERROR (WrapperStartStopAppMain) [Main:119] Error encountered during MID server execution
java.lang.NullPointerException: Cannot invoke "org.w3c.dom.NodeList.getLength()" because "nl" is null"

The two characters that I found that will certainly result in a failure to start the service are '<' and '&'; Once the password was updated to exclude any problematic special characters, the MID Server service was able to start and our MID was finally able to validate.

TLDR; If you are facing issues with your MID eternally validating post-clone, make sure that your MID Server User's password matches what is found in the config file on the server, as well as verifying that there are no special XML characters (< or &) in the password.