Since Wednesday morning we have experienced problems with the EPP service and the registry database has been slower than usual.

In May when we introduced a new data model in the registry system, many objects were deleted from the database. Yesterday morning these objects were removed, and this generated more service messages than normal. We now see that there is a weakness in how the system handles a long queue of service messages.

In order to solve the problems today, we changed all service messages of the type"object-deleted" to read. This caused the load on the database server to go down to a normal level. The system is now running as normal.

We apologise for the problems this has caused for our registrars.

Published: 5 July 2018