Playback is where the MongoDB blog brings you selected talks from around the world and around the industry. Here, we continue show-casing the great talks at MongoDB World 2018 from people running MongoDB in production and at scale.
Travel technology company Amadeus IT Group run some 100 MongoDB clusters with around 3000
mongod nodes. Some of those clusters have over 120 shards alone and at that scale, you need all the automation you can get your hands on. For Arkadiusz Borucki, Senior Service Reliability Engineer at Amadeus, that means using MongoDB Ops Manager as part of a completely automated environment. In his talk at MongoDB World 2018, he takes us through the range of technologies involved in that environment.
It starts with four MongoDB Ops Manager instances, three directly managing instances and one master. Under their watchful eye, there's a wide range of different MongoDB use cases - search, analytics, payments, fraud detection and more - that need a range of different services. The solution involves writing code to define infrastructure, storing that code in git, getting Jenkins to act on changes and running Ansible to orchestrate the MongoDB changes through Ops Manager.
In the course of the talk, Borucki shares examples of the code that drives the process, leveraging Ops Manager's REST APIs, and looks at the code Amadeus has open sourced to complement Ansible's functionality.