DDB Async architecture options

options.md

• Introduce a caching layer.
  • Do a query to call a Lambda. That lambda calls another lambda in async mode and the 1st lambda responds a req id to caller
  • The 2nd lambda queries ddb and waits for result. On result received from ddb write results to a cache like Elasticache
  • From frontend keep querying cache with request id. WHen result gets available, read the results. This will be much faster than ddb
• Use websocket
  • Implement a websocket service
  • Open a websocket connection from frontend
  • repeat above steps with 2 lambda (from caching option)
  • but instead of cache, 2nd lambda returns results to websocket and frontend gets result from socket connection
• Use DAX to improve performance. This will improve DDB performance if you have large number of records
• Use multiple DDB tables to categorize data. Query specific tables based on what you want to query. This may not apply on all scenarios and very use case specific