Implement Query Side Optimization - The Reactor

Implementation

The query side is the read path in CQRS. Instead of querying the write model directly (which is normalized for writes), it reads from pre-built read models (projections) that are denormalized and indexed for the specific query being served. Query results can also be cached to avoid redundant work.

Implement a node that serves three query types:

// Paginated user list from a denormalized listing projection
{ "type": "GetUserListing", "msg_id": 1,
  "params": {"page": 1, "limit": 10} }
-> { "type": "query_result", "in_reply_to": 1,
    "data": [{"id": "user-123", "name": "John Doe"}],
    "cached": false }

// Email lookup via index (second call returns from cache)
{ "type": "GetUserByEmail", "msg_id": 2,
  "params": {"email": "john@example.com"} }
-> { "type": "query_result", "in_reply_to": 2,
    "data": {"email": "john@example.com", "userId": "user-123"},
    "cached": true }

// Filtered query: users in a specific city
{ "type": "GetUsersByCity", "msg_id": 3,
  "params": {"city": "NYC"} }
-> { "type": "query_result", "in_reply_to": 3,
    "data": [{"city": "NYC", "userId": "user-123"}],
    "cached": false }

The cached field indicates whether the result was served from the query cache. A query handler never modifies any state.

Sample Test Cases

Query user listingTimeout: 5000ms

Input

{
  "src": "client",
  "dest": "queryside",
  "body": {
    "type": "GetUserListing",
    "msg_id": 1,
    "params": {
      "page": 1,
      "limit": 10
    }
  }
}

Expected Output

{"type": "query_result", "in_reply_to": 1, "data": [{"id": "user-123", "name": "John Doe"}], "cached": false}

Query with cache hitTimeout: 5000ms

Input

{
  "src": "client",
  "dest": "queryside",
  "body": {
    "type": "GetUserByEmail",
    "msg_id": 1,
    "params": {
      "email": "john@example.com"
    }
  }
}

Expected Output

{"type": "query_result", "in_reply_to": 1, "data": {"email": "john@example.com", "userId": "user-123"}, "cached": true}

Hints

Hint 1▾

Read models are denormalized: pre-join and pre-aggregate data for fast reads

Hint 2▾

Cache query results with a TTL; return cached=true when the result comes from cache

Hint 3▾

GetUserListing uses pagination (page, limit) to return a slice of the users list

Hint 4▾

GetUserByEmail uses an email index for O(1) lookup rather than a full scan

Hint 5▾

The query side never writes — it reads from projections built by the event side

Implementation

Sample Test Cases

Hints

Resources

Theoretical Hub

Key Concepts