drove cluster¶

Manage and monitor your Drove cluster.

Synopsis¶

drove cluster <command> [options]

Commands¶

Command	Description
`ping`	Check cluster connectivity
`summary`	Show cluster resources
`leader`	Show current leader
`endpoints`	List exposed endpoints
`events`	Stream cluster events
`simulate-placement`	Simulate placement for a deployable spec
`maintenance-on`	Enable maintenance mode
`maintenance-off`	Disable maintenance mode

drove cluster ping¶

Check whether the cluster is reachable.

drove cluster ping

drove cluster summary¶

Show how much of the cluster resources are in use.

drove cluster summary

Output:

State                         NORMAL
Leader Controller             drove-ctrl-1:4000
Cores                         Utilization: 45% (Total: 256 Used: 115 Free: 141)
Memory                        Utilization: 62% (Total: 512,000 MB Used: 317,440 MB Free: 194,560 MB)
Number of live executors      8
Applications                  Active: 23 Total: 31

drove cluster leader¶

Show the current cluster leader.

drove cluster leader

drove cluster endpoints¶

List all exposed application endpoints. Use this when you debug routing.

drove cluster endpoints [--vhost VHOST]

Option	Description
`-v, --vhost`	Filter by virtual host

Example:

drove cluster endpoints -v myapp.example.com

drove cluster events¶

Stream cluster events in real time.

drove cluster events [--follow] [--type TYPE] [--count COUNT]

Option	Description
`-f, --follow`	Stream continuously
`-t, --type`	Filter by event type
`-c, --count`	Number of events to fetch at a time (default: 1024)
`-s, --textfmt`	Python format string for each event. The default prints the event type, id, time, and metadata, separated by pipes.

Example:

# Watch all events
drove cluster events -f

# Filter by type
drove cluster events -t APP_STATE_CHANGE -c 50

drove cluster simulate-placement¶

Simulate placement for a deployable spec. The command sends the spec to the cluster simulation API, and the cluster reports where the instances would run. The cluster does not change.

drove cluster simulate-placement SPEC_FILE [--num-instances N] [--placement-policy FILE] [--detail] [--json]

Option	Description
`spec_file`	Path to a spec file. The spec can be an application, task, or local service.
`-n, --num-instances`	Number of instances to simulate. The default is 1.
`--placement-policy`	Path to a JSON file with a placement policy. The policy overrides the spec policy.
`--detail`	Show per-instance placements.
`--json`	Print the raw JSON response.

Example:

drove cluster simulate-placement app-spec.json -n 4 --detail

drove cluster maintenance-on¶

Put the cluster in maintenance mode. The cluster rejects new operations.

drove cluster maintenance-on

Warning

Running applications continue but no new deployments or scaling operations are accepted.

drove cluster maintenance-off¶

Take the cluster out of maintenance mode.

drove cluster maintenance-off