Cache Demo

a demo for cache and use postgreSQL as communication latency

Explanation

The core design of our cache implementation is located in cache.py.

I use the Least Recently Used (LRU) algorithm as the cache eviction policy.

I use two local variables:

1. lru_cache: This stores the actual cached data following the LRU policy.
2. temp_buffer: This temporarily holds keys that are encountered for the first time.

• When a key is accessed for the first time, it is added to temp_buffer means an initial counter value of 1.
• If the key is accessed again (i.e., it's a "being used"), the key is removed from temp_buffer and inserted into the head of lru_cache with its associated value.

due to

5. Only consecutive cache hits (2 or more times) are counted as "being used."

so, the temp_buffer size set to 1. (That can be extend to more if needed)

This mechanism ensures that only keys with consecutive hits are promoted to the main cache.

and this work is using ChatGPT to support coding and code review, this is the chat history

Run the scripts:

I use PostgreSQL to sim internal communication and disk fetching for measure speed.

• env build

  docker compose build
  docker compose up -d
  docker compose exec py2db python library_create.py
  docker compose exec py2db python library_add.py
  

• running cache demo

  docker compose exec py2db python demo_cache.py
  

compare run 100 time of without cache vs with cahce: