This is a Ruby client and a command-line tool for the MongoDB Atlas API . Other Cloud APIs may be added in the future based on demand.
This tool is an alternative to mongocli. It was started to address the following goals:
- Support for actions that mongocli did not implement (such as Atlas organization management).
- Easier to use command line argument/option structure.
- A Ruby API client consumable by other libraries/programs.
Credentials can be given as top-level command-line options:
cloud -u $USERNAME -p $PASSWORD ...
They can also be supplied via mongocli-compatible environment variables:
export MCLI_PUBLIC_API_KEY=$USERNAME MCLI_PRIVATE_API_KEY=$PASSWORD
To use atlas-dev, set MCLI_OPS_MANAGER_URL:
export MCLI_OPS_MANAGER_URL=https://cloud-dev.mongodb.com
The Atlas API generally requires multiple parameters to identify any given resource, and there is no consistent scheme that is used for resource identification. For example, to operate on clusters one must supply a project id and a cluster name.
mongo-cloud by default assumes that all resources specified by the user are identified by their id. For example, to view a cluster, one might issue:
mongo-cloud cluster -p PROJECT-ID -c CLUSTER-ID show
The cluster can also be identified by its name, for compatibility with other tools:
mongo-cloud cluster -p PROJECT-ID --cluster-name CLUSTER-NAME show
For convenience, whenever a resource identifier is needed, mongo-cloud
accepts both the id and the customary identifier used by Atlas API for the
resource in question. For clusters this means that cluster name can be used
as the parameter to the -c argument:
mongo-cloud cluster -p PROJECT-ID -c CLUSTER-NAME show
Note that mongo-cloud will first attempt to interpret each identifier as the id even if the underlying API endpoint specifies a different field such as the name.
Since providing multiple identifiers is in most cases redundant, mongo-cloud transparently stores mappings of both id mappings and resource parent mappings when resources are retrieved. For example, when retrieving a list of clusters in a project, mongo-cloud stores the mapping from cluster ids to cluster names and the mapping from cluster ids to project ids for each cluster. Subsequently a cluster can be viewed by specifying only the cluster id:
mongo-cloud cluster -p PROJECT-ID list
mongo-cloud cluster -c CLUSTER-NAME show
Organization:
mongo-cloud org list
mongo-cloud org show 1234cafe...
Projects:
mongo-cloud project create --org ORG-ID --name PROJECT-NAME
mongo-cloud project list
mongo-cloud project show 1234cafe...
IP whitelists:
mongo-cloud whitelist -p 1234cafe... list
mongo-cloud whitelist -p 1234cafe... show 0.0.0.0/0
mongo-cloud whitelist -p 1234cafe... add 0.0.0.0
Database users:
mongo-cloud dbuser -p 1234cafe... list
mongo-cloud dbuser -p 1234cafe... show UserName
mongo-cloud dbuser -p 1234cafe... create UserName Password
Clusters:
mongo-cloud cluster -p PROJECT-ID create --name NAME --config YAML-CONFIG
mongo-cloud cluster -p PROJECT-ID list
mongo-cloud cluster -p PROJECT-ID show ClusterName
mongo-cloud cluster [-p PROJECT-ID] -c CLUSTER-ID reboot
MIT