cache types: how do I identify error types?

[smousa]

Hello,

When building a readthrough cache in a pipeline, is there a way I can distinguish from an error where I get a cache miss vs some other type of error?

processor_resources:
  - label: get
    processors:
      - try:
        - cache:
            resource: my_cache
            operator: get
            key: MYKEY
      - catch: # how do identify what I am catching is no records found or something is wrong with the cache?
        - branch:
            result_map: root.value = this
            processors:
              - resource: call_database
        - branch:
            result_map: root = this.value
            processors:
              - try:
                - cache:
                    resource: my_cache
                    operator: add
                    key: MY_KEY
                    value: this.value

[mihaitodor]

Hey @smousa 👋 You can use a switch processor and check the suffix of error(), something like check: error().has_suffix("key does not exist") to detect if it was a cache miss or, depending on the cache type, some other error. Benthos will surface connection errors from various caches, but it doesn't attempt to expose generic errors. For example for aws_s3, you might get something like "operator failed for key 'test': operation error S3: GetObject, resolve auth scheme: resolve endpoint: endpoint rule error, Invalid region: region was not a valid DNS name." or for redis you might get "operator failed for key 'test': dial tcp [::1]:6397: connect: connection refused". However, I think you only care if it's a cache miss or not.