InMemroy cache of schema resulting in un-expected behavior
Schemas are cached in memory of the running application. Because of this inmemory caching the following scenario will fail. Let us assume we have 2 instances of application running I1 and I2.
- Create schema - Lands on I1
- Get schema - Lands on I1. This schema is cached on I1
- Delete schema - Lands on I2. This will try to delete any cache entry for that schema in I2. So far the cache entry on I1 is still intact.
- Get schema - Lands on I1. As the cache entry is not cleared it will return 200 instead of 404 not found.
The below test along with other tests are failing intermittently because of the above mentioned issue.
- should_createSchema_and_returnHttp409IfTryToCreateItAgain_and_getSchema_and_deleteSchema_when_providingValidSchemaInfo
If the service is running on multiple VMs or Pods, this issue will be prominent and will block pipelines and results in un-expected behavior