In 11.0.9.version we released an improvement for memory handling to be able to scale workload in a more efficient way. This was tested in our test environments but no issues were found.
After releasing 11.0.9. we monitored the solution and noticed that the application servers were not performing as expected and processes were taking too long to complete and were timing out.
Very soon we concluded on the root cause and decided to release a patch version which fixed the issue.
We will add a mechanism to prevent processes to run too long on the servers and to prevent integrations to trigger the same process again in case we have a similar issue. This will stop the problem from escalating.
We have also agreed on a process that should be followed with similar changes so that we are more aware of the risk and are able to notice it earlier in the development process.