Custom serialization

One of the joys of using Rama is being able to use first-class types all the time, whether appending data to a depot, processing data in ETLs, querying PStates, or anything else. To use an object type within Rama, Rama must know how to serialize/deserialize that type to a stream of bytes. Rama does this when writing objects to disk or transferring objects to other processes.

Rama has built-in support for using basic types (int, long, float, double, string, etc.), commonly used types from java.util, Clojure data structures, and types defined with defrecord.

For other types you need to tell Rama how to serialize/deserialize them. One option is to use the serialization mechanisms described on the main serialization page, such as implementing the RamaCustomSerialization interface.

Under the hood, Rama uses Nippy for serialization. So you can also register serializations with Rama by extending Nippy’s protocols directly.

Something to watch out for with Nippy is it uses a 2 byte hash to encode type IDs when using extend-freeze. This makes collisions of type IDs occur with greater than 1% probability after only 50 extensions (birthday problem). Rama’s RamaCustomSerialization interface implements Nippy extensions in a special way to internally use an 8 byte hash of those types to make collisions effectively impossible.