Previous incidents

One of our two backends had issues with caches. The issue resolved automatically after 30 minutes.
We are still investigating the root cause for future prevention