Tell us what failed most recently, which applications were involved and how someone put it right. That gives us a starting record to trace through the code, settings and logs.
An investigation before a fix
The cause might be a changed interface, a response the code does not handle or a business rule omitted from the original build. We explain the evidence and proposed repair before changing the connection.
We test the failed case alongside normal transfers, check whether existing records need attention and update the recovery instructions. If replacement is the better route, we explain what would carry over.
Describe the recent failure. Please leave credentials and customer records out of the first message.