Cache related service degradation. (Flows stop working, updates start failing with "Maximum call stack size exceeded") about directus

[br41nslug]

Feels related to #19664

[rijkvanzanten]

@a14e How much memory do your Directus containers have? Gut feeling says it's the combination of having a somewhat larger schema, a lower memory ceiling for NodeJS, not using Redis for caches, and our default configuration of 100 cached GraphQL schemas. Lowering the GRAPHQL_SCHEMA_CACHE_CAPACITY to fit your schema and memory configuration is actually the correct solution here, unless one of my assumptions isn't accurate 🙂

Closing this for now as the presented workaround is the correct solution. Happy to reopen and keep discussing if other information becomes available

[a14e]

@rijkvanzanten I'm using Redis cache. I have 1 GB of RAM. I tried to increase it to 2 GB, but the situation remained the same. There is also a memory usage plot in my initial report. Memory usage was under 200 MB.