TTL for empty results in lookup tables

[patrickmann]

Resolves #16466

Description

Adds optional TTL for empty results in lookup table caching. This is not available if ignore null has been checked.

Motivation and Context

If we are caching empty results, then by default they live forever. This is not compatible with data adapters where the value may change over time, specifically DNS. A fixed expiration (expire after write) is not helpful, because that would potentially expire non-null values prematurely. By tuning the TTL for empty results we can avoid requerying the data source excessively, while still becoming aware of changes in values.

How Has This Been Tested?

Simple test scenario:

• create a CSV file lookup table:
  -- data adapter check interval = 10s
  -- cache: empty TTL = 30s and ignore null = false)
• query for a non-existent key x: result null
• update CSV to include x
• requery before 30s have elapsed: result null
• requery after 30s: result is the newly saved value from CSV file

Empty TTL overrides the expire after access setting (for null values). Otherwise we would never update a cached null value, that is being requeried very frequently.

Screenshots (if appropriate):

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Refactoring (non-breaking change)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• My code follows the code style of this project.
• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• I have read the CONTRIBUTING document.
• I have added tests to cover my changes.

[kodjo-anipah]

LGTM,
Tested with csv.

1. key does not exist tll empty not configured -> never updates
2. key does not exist first ttl empty configured, key added later -> updates
3. ignore empty results -> configuring value for ttl empty is deactivated