Skip to content

Latest commit

 

History

History
145 lines (105 loc) · 5.89 KB

infrastructure.md

File metadata and controls

145 lines (105 loc) · 5.89 KB
Raw

joynr infrastructure

In order for joynr participants to be able to communicate with each other, a few infrastructure components are necessary for them to be able to discover each other and correctly route messages between the participants.

This documents describes some examples of setting up these infrastructure components.

Components

The following component is necessary in order to set up a joynr environment:

  • Global Discovery service (GlobalCapabilitiesDirectory)
    • A joynr based application with which the participants register their providers and query to discover other participants

Additionally, you have to ensure that any required transport layer components, such as MQTT brokers, are set up.

Java

This section describes configuring and starting the JAVA version of the Discovery service.

GlobalCapabilitiesDirectory: see GCD Readme.

JEE

This section describes configuring and installing the JEE versions of the Discovery service in a Glassfish / Payara container. This document assumes you are familiar with the basics of installing and setting up the container itself.

Please note that as two applications use the joynr JEE integration layer, all the documentation in the JEE Developer Guide is relevant here, also. Specifically, setting up the managed scheduler executor service is mandatory.

The documentation assumes you have already installed Glassfish / Payara and will reference the root of the installation as ${GF_HOME}.

Creating the domain and an executor service

A recommended first step is to copy the the domain1 directory to create a fresh domain for the joynr infrastructure:

cd ${GF_HOME}/glassfish/domains
cp -R domain1 joynr-infrastructure
cd joynr-infrastucture/config
sed -i 's/domain1/joynr-infrastructure/' domain.xml

Next, we start the new domain and create the executor service:

cd ${GF_HOME}
bin/asadmin start-domain joynr-infrastructure
bin/asadmin create-managed-scheduled-executor-service --corepoolsize=100 concurrent/joynrMessagingScheduledExecutor

Note the --corepoolsize=100 option. The default will only create one thread, which can lead to blocking.

Database

The Discovery service requires a database to store its data. Which database you use is principally up to you, but you may have to customise the schema creation if the standard JPA mechanisms don't work with your database of choice.

This document will describe setting up the Derby database service which is shipped with Glassfish / Payara.

cd ${GF_HOME}
bin/asadmin create-jdbc-connection-pool --datasourceclassname org.apache.derby.jdbc.ClientDataSource --restype javax.sql.XADataSource --property portNumber=1527:password=APP:user=APP:serverName=localhost:databaseName=joynr-discovery-directory:connectionAttributes=\;create\\=true JoynrPool
bin/asadmin create-jdbc-resource --connectionpoolid JoynrPool joynr/DiscoveryDirectoryDS
bin/asadmin start-database

The last command starts the local Derby database instance. This required before deploying the applications, and also necesary before starting the applications. Remember to start the database each time before starting the application server subsequently.

Other databases

In a production environment, you most likely won't want to use the Derby database, but instead an enterprise level database such as PostgreSQL, Oracle, DB2, etc. In that case, note that you might have to alter the persistence.xml files of the project Discovery Directory JEE in order to match your database.

You also might not want to use the automatic JPA schema creation in that case, and instead manually maintain the schema or use something like Flyway. Both of these issues are outside the scope of this guide and are not further documented here.

Of course, the asadmin command documented above for creating the JDBC connection pool will also not be relevant when using a different database system. See the Glassfish / Payara documentation on how to set up the connection pool for your database. The commands for creating the data sources for the connection pool should be re-usable as is, assuming you named your connection pool JoynrPool as above.

Deploying the applications

This guide assumes you have checked out the joynr source code and have built the entire project, so that the build artifacts for the Discovery service is present on your local drive.

${JOYNR_HOME} is used to denote the root directory to which you checked out the source code. ${JOYNR_VERSION} is used to denote the version of joynr you have checked out and which appears as part of the filename for the WAR files created by the build.

Single backend (default)

cd ${GF_HOME}
bin/asadmin deploy ${JOYNR_HOME}/java/backend-services/discovery-directory-jee/target/discovery-directory-jee-${JOYNR_VERSION}.war

Multiple backends

See multiple backends guide for details about the multiple backends feature.

You can find an example for setting up a JEE based joynr infrastructure for two backends as Docker images as part of the system integration tests:

Logging

If you need to see more detail about communication, or if you are experiencing problems starting up the applications and want more details about what is going wrong, then you can set the log levels as follows:

cd ${GF_HOME}
bin/asadmin set-log-levels io.joynr=FINE

Of course, you can be more specific about which joynr packages to log at which level. Also, you can switch logging to FINEST for even more log details.