Java module to analyze cache hit ratio from incoming search logs
- Parses logs and generates unique hash using murmur hash for every incoming query (q-param from the log)
- Constructs a final data structure based on the cache TTL value set in the CacheAnalysis.java and finally writes the data structure contents to a flat file
- Running the attached awk script on the final flat file tells you how much of your cache is really used
- Do necessary customizations(cache ttl, incoming query param name etc) to the CacheAnalysis.java
- Run runcache.sh
- Finally run awkscript.txt