Storage based on probabilistic structure Bloom filter
Suitable for storing huge amount of data and quering if element is present or not
You probably should use https://github.com/tivvit/aerospike-pure-lua-bloom-filter which uses Aerospike bytes directly
This module is based on: https://github.com/mozilla-services/lua_bloom_filter but it uses specific serialization methods (https://github.com/mozilla-services/lua_bloom_filter/pull/4)
This module uses Aerospike's bytes.get_string() which causes memory leaks (aerospike/aerospike-mod-lua#3)
- Compile lua_bloom_filter (described in the project)
- Upload so module to Aerospike
echo "REGISTER MODULE '/path_to_repo/lua_bloom_filter/release/bloom_filter.so'" | aql - Upload UDF to Aerospike
echo "REGISTER MODULE '/path_to_repo/bloom.lua'" | aql
Use Aerospike apply
Module: bloom
Method: add
Example in python:
as_client.apply(key, "bloom", "add", ["bin", value])) #returns 0 if not found, 1 if foundDefault size is 10000 elements with precision 0.01, which are defined on first lines of the code. I did not want to pass them as parameter with every request, if you have better idea how to solve that fell free to open issue or PR.
Feel free to contribute.
© 2016 Vít Listík
Released under MIT licence