Database connection, jvm

To use Molecule on the server/jvm side we need a few imports and a connection to a Datomic database.


Import the Molecule api and the generated DSL for your Data Model

import molecule.datomic.api._
import app.dsl.yourDomain._

Also, choose a Datomic database system:


The Datomic Peer runs in your application process.

import molecule.datomic.peer.facade.Datomic_Peer._

Peer Server

The Datomic Peer Server runs on a remote server (could be local also).

import molecule.datomic.client.facade.Datomic_PeerServer._

Dev-local (Cloud)

Datomic dev-local provides a local testing environment for Datomic Cloud without connecting to a server.

import molecule.datomic.client.facade.Datomic_DevLocal._


Get a connection to the database in one of 3 ways:

Re-create database and schema

import app.schema.YourDomainSchema

// In-memory connection (default) 
implicit val conn = recreateDbFrom(YourDomainSchema)

// .. or with storage service
// WARNING: this completely destroys the database and creates a new empty one!
implicit val conn = recreateDbFrom(YourDomainSchema, datomic-db-uri, protocol)

Transact schema (migrate)

import app.schema.YourDomainSchema

// In-memory connection (default) 
implicit val conn = transactSchema(YourDomainSchema)

// .. or with storage service
implicit val conn = transactSchema(YourDomainSchema, datomic-db-uri, protocol)

We can transact our complete schema as often as we want. If a transaction value is the same as the current value in the database, Datomic simply ignores it. Whereas changes we make to our Data Model / Schema will be transacted.

This makes it easy to migrate our schema: we make changes to our Data Model, sbt compile -Dmolecule=true and transact our updated generated Schema transaction file.

And if some of our code is still using an old attribute definition, the compiler will warn us. The compiler in this way help us enforce that our code, data, schema and Data Model all stay in sync! We won’t be able to save data with a molecule that is outdated since it won’t compile.


When no changes are needed to our schema, we simply connect to our database.

implicit val conn = connect(datomic-db-uri, protocol)

Datomic db URI + protocol

Datomic uses a URI to configure the connection to its database systems and describing what underlying Storage Service is used in this basic form:


The protocol is “mem” for an in-memory database, “free” (Free)/“dev” (Starter/Pro) for a development database saving to local disk, “cass” for an underlying Cassandra disk storage system and so on.

The db-identifiers for an in-memory database could be “hello”, and the complete uri would be “datomic:mem://hello”.

With a transactor running and a local Datomic database saving to disk, the identifier could be “localhost:4334/mbrainz-1968-1973” which tells Datomic how the database is reached through localhost on port 4334 and that the database name is “mbrainz-1968-1973”.


Setting up a Client database connection