Builds out a Jenkins build-follower for SQLAlchemy and related projects.
The Jenkins that runs at http://jenkins.sqlalchemy.org is only the front end jenkins servers and no jobs are actually run. The jobs are run on remote build-followers that run in a private network. These machines call-in to the Jenkins server that's exposed publicly.
A build follower basically contains lots of Python builds as well as running instances of all the major databases we test on. The databases are running in Docker containers:
[classic@dell ~]$ sudo docker ps --format "table {{.ID}}\t{{.Image}}\t{{.Ports}}\t{{.Names}}"
CONTAINER ID IMAGE PORTS NAMES
af61f4b6d9f4 mssql_2017 192.168.1.208:1433->1433/tcp mssql_2017
2a0ffcfb23bb mariadb_10.1.24 192.168.1.204:3306->3306/tcp mariadb_10.1.24
c0548ccc0f18 mariadb_10.2.9 192.168.1.203:3306->3306/tcp mariadb_10.2.9
d8bd8a83cde8 mysql_5.6 192.168.1.202:3306->3306/tcp mysql_5.6
00426728f505 mysql_5.7 192.168.1.201:3306->3306/tcp mysql_5.7
e024673df0b6 postgresql_9.4 192.168.1.207:5432->5432/tcp postgresql_9.4
c48927ea483d postgresql_9.6 192.168.1.206:5432->5432/tcp postgresql_9.6
36dc50aec4d3 postgresql_10 192.168.1.209:5432->5432/tcp postgresql_10
0ccc4cc4bcf9 oracle_11.2.0 192.168.1.205:1521->1521/tcp, 192.168.1.205:8085->8085/tcp oracle_11.2.0
The Docker containers are all custom and work the same way. The Oracle and MSSQL containers are the most elaborate.