Recap of issue: In 3.11.0 we will be deleting a failed namespace to reduce potential for spamming a cluster and creating hundreds of namespaces on failed deprovisions. In 3.11.z we would like to go back to previous behavior of keeping transient namespace around on failures to help aid debugging. This BZ is tracking work to address how we can handle service-catalog calling the broker in a loop on failed deprovisions.
Service Catalog bz possibly related Bug 1628235 - Service Catalog establishes a new connection for every request and never closes them.
This bug is fixed by deleting the failed namespaces and to go back to the previous behavior is an RFE being tracked in the duplicate bug. Closing in favor of more correct bug title. *** This bug has been marked as a duplicate of bug 1649076 ***