AutoPilot Pattern MongoDB

A robust and highly-scalable implementation of MongoDB in Docker using the Autopilot Pattern

Architecture

A running cluster includes the following components:

• ContainerPilot: included in our MongoDB containers to orchestrate bootstrap behavior and coordinate replica joining using keys and checks stored in Consul in the health, and onChange handlers
• MongoDB: we're using MongoDB 3.4 and setting up a replica set
• Consul: used to coordinate replication and failover

Running the cluster

Starting a new cluster is easy once you have your _env file set with the configuration details

• for Triton, just run docker-compose up -d
• for non-Triton, just run docker-compose -f local-compose.yml up -d

In a few moments you'll have a running MongoDB ready for a replica set. Both the master and replicas are described as a single docker-compose service. During startup, ContainerPilot will ask Consul if an existing master has been created. If not, the node will initialize as a new MongoDB replica set and all future nodes will be added to the replica set by the current master. All master election is handled by MongoDB itself and the result is cached in Consul.

Run docker-compose -f local-compose.yml scale mongodb=2 to add a replica (or more than one!). The replicas will automatically be added to the replica set on the master and will register themselves in Consul as replicas once they're ready.

Configuration

Pass these variables via an _env file.

• LOG_LEVEL: control the amount of logging from ContainerPilot
• when the primary node is sent a SIGTERM it will step down as primary; the following control those timeouts
  • MONGO_SECONDARY_CATCHUP_PERIOD: the number of seconds that the mongod will wait for an electable secondary to catch up to the primary
  • MONGO_STEPDOWN_TIME: the number of seconds to step down the primary, during which time the stepdown member is ineligible for becoming primary
  • MONGO_ELECTION_TIMEOUT: after the primary steps down, the amount a tries to check that a new primary has been elected before the node shuts down
• CONSUL (optional): when using local-compose.yml, this will default to consul (and thus use the DNS provided by Docker), but for deploying on Triton via docker-compose.yml, this should be set to the CNS path of the consul service (consul.svc.XXX...)

Not yet implemented:

• MANTA_URL: the full Manta endpoint URL. (ex. https://us-east.manta.joyent.com)
• MANTA_USER: the Manta account name.
• MANTA_SUBUSER: the Manta subuser account name, if any.
• MANTA_ROLE: the Manta role name, if any.
• MANTA_KEY_ID: the MD5-format ssh key id for the Manta account/subuser (ex. 1a:b8:30:2e:57:ce:59:1d:16:f6:19:97:f2:60:2b:3d); the included setup.sh will encode this automatically
• MANTA_PRIVATE_KEY: the private ssh key for the Manta account/subuser; the included setup.sh will encode this automatically
• MANTA_BUCKET: the path on Manta where backups will be stored. (ex. /myaccount/stor/triton-mysql); the bucket must already exist and be writeable by the MANTA_USER/MANTA_PRIVATE_KEY

Sponsors

Initial development of this project was sponsored by Joyent.