The Database On Demand platform provides the CERN user community with an automatized solution for operations such as database creation,
configuration and destruction; as well as mission critical procedures like
database backup and recovery operations or database system upgrades.
The CERN Database On Demand service manages over 530 (at the time of this
submission) databases from different vendors (PostgreSQL, MySQL, InfluxDB and Oracle), supporting many different use cases for the CERN user community:
sensor data, administrative databases, metrics logging, infrastructure system
back ends, etc.
This session would provide an overview of the Database on Demand platform and service functionalities, explaining how we profit from Puppet in conjunction
with our Open Source DB On Demand API component acting as a source of configuration information to reduce the impact of operations on the Database team during the different stages in the life cycle of a database system, including taking care of database systems initialization, the initial database accounts management, configuration and registration of databases in the monitoring system configuration (created using Open Source tools like InfluxDB, ElasticSearch and Grafana) plus other operations such as configuration of second level backups or database instances migration between hosts.