Experimenting with lagom-java, custom deploy (no conductR)

Using Maven.

Currently not using any PersistentEntities, but using akka cluster directly. Using cluster sharding backed by Akka Distributed Data.

Using straight up old school jdbc to update a postgres db.

Simulating a non conductR build

JDBC

I Used PGAdmin to do the following

* Install postgres
* Create a db called "playdb"
* Create a login/role called "playuser" with passwword "playuser"
* Grant this role permissions to playdb


In saga-impl/application.conf the following is used for db connection:

db.default {
  driver = "org.postgresql.Driver"

  #jdbc:postgresql://localhost/playdb is fine... mine runs on 5433, so I put it there, usually its 5432
  url = "jdbc:postgresql://localhost:5433/playdb"

  username = playuser
  password = playuser
}

Cluster

The recommendation is to have an akka cluster per service (NOT leveraging cluster for service discovery, ie. multiple
services in one cluster).


The following in saga-impl/application.conf set this up:

akka{
  remote {
    netty.tcp {
      hostname = "127.0.0.1"
    }
  }
}

and in saga-impl/start.sh

CONFIG="-Dhttps.port=9443 -Dplay.crypto.secret=$PLAY_SECRET  -Dakka.cluster.seed-nodes.0=akka.tcp://application@127.0.0.1:2552 -Dakka.remote.netty.tcp.port=2552"

java -cp "target/saga-v1/saga/lib/*" $JAVA_OPTS $CONFIG play.core.server.ProdServerStart

Maven package and run

start first node from root dir:

sh saga-impl/start.sh (zip, unzip, and run).

start second node:

sh saga-impl/start2.sh

debugging in intellij

start server in debug mode on 5555:

env MAVEN_OPTS="-Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=y,address=5555" mvn lagom:runAll

create a maven remote debug configuration in intellij, attaching to 5555