Hi, this has already been raised with DCP173217 but I am also putting on here as suggested.

We have elements set up in a DMA called awsdmt501a that contains an API layer so that it can receive http calls from outside Dataminer, which works fine when the calls are made directly to a DMA (awsdmt501a in this case). However, after we rolled out a new certificate on awsdmt501 for the new shared hostname, only 50% of the calls made to that shared hostname reached the elements which means that the other 50% is trying to hit 501b instead of 501a and got lost, which is unexpected.

Accessing to the cube through the new shared hostname works fine and below is the current failover setting which indicates all the calls made to ‘awsdmt501’ should be directed to ‘awsdmt501a’.