I have a very general query based on its filter performance.
I have 30 million records of data where I need to do 5 filters on an average per query to get the records on each set.
I am using graph database and aerospike together. I use graph-database for geo-based filtering as I could not find anything on aerospike, but my graph database is horrible, horribly, horribly slower with multiple filters and I am looking for an alternative.
Though I know aerospike is faster in reading data, but I want to know from other people on how aerospike performs with multiple filters like (3-5 filters)?
Usually this is done faster by writing reference data into another key. If your use case is geo-based filtering, are you using https://www.aerospike.com/docs/guide/geospatial.html ? How are you filtering currently? Other than using reference keys I think the other way you could do it is with predicate fitlers, but it really depends on how fast you need your transaction to be…
For geo query, I am using graph database from another vendor - dgraph . But For multiple filter in query, it is very very slow. So, I am looking for an alternative in Aerospike.
Whether aerospike returns data with 3-5 filter as fast as it returns other query (& faster than my current dgraph choice)
Multiple Filter like -
type = "foo" AND category= "bar" AND tags = "baz"
with higher performance