An overall slowness and errors affected one server of our On Demand application.
Users directed by the load balancer to the affected server experienced slowness, disconnections, and errors.
[17/01/2019, 06:30 AM EST] A huge pike of wrong SQL requests caused slowness on the affected server, during about one hour. We identified what method was called, and protected the code to prevent those requests to happen again. We're still investigating how the method could have been called so many times with wrong parameters. We set a logger to track any further issue.
[16/01/2019, 11:09 AM EST] Problem fixed, server stabilized, origin of the problem still being investigated