Configuration

coco’s main configuration file has the following options. It can be passed to coco with coco[d] -c path/to/coco.conf.

log_level : str

The global log level. Can be one of CRITICAL, ERROR, WARNING, INFO or DEBUG. Default INFO.

host: str

Host the coco server (cocod) is run on.

port: int

Port the coco server (cocod) should be run on. Default 12055.

metrics_port: int

Port cocod should run its metrics server on. Default 9090.

endpoint_dir: str

Path where endpoint config files are located.

n_workers: int

Number of sanic workers to start for the frontend of cocod. Default 1.

session_limit: int

Maximum number of tasks being executed concurrently by request forwarder. A higher number will use more memory. Default 1000.

blocklist_path: str

Path to persistent blocklist storage file. Default /var/lib/coco/blocklist.json.

storage_path: str

Path to persistent state storage. Default /var/lib/coco/state/.

groups:

Groups of nodes that are managed by coco. Contains key-value pairs where the key is the name of the group (str) and the value is a list of str in the format host:port.

Example:

groups:
    gps_server:
        - carillon.chime:54321
    cluster:
        - localhost:12050
        - localhost:12000
    receiver_nodes:
        - recv1:12048
        - recv2:12048
    all:
        - localhost:12050
        - localhost:12000
        - recv1:12048
        - recv2:12048

load_state:

Initial cluster state. Contains key-value pairs that are state paths and file system paths. The files located at the latter locations are loaded into the prior parts of the state when resetting the state. Default: None.

Example:

load_state:
    cluster: "../conf/gpu.yaml"
    receiver: "../conf/recv.yaml"

slack_token: str

Slack authorization token. Default: None.

slack_rules: list

Rules for dispatching logging messages to slack. These specify the logger path, the minimum level it applies to and the slack channel the messages should go to. Default: None.

Example:

slack_rules:

    - logger: coco
      level: WARNING
      channel: coco-alerts

    - logger: coco.endpoint.update-tracking-pointing-0
      level: INFO
      channel: pulsar-timing-ops

queue_length: int

Length of the endpoint request queue between front- end backend, managed by redis. Default 0.

timeout: str

Time before requests sent to nodes time out. A string representing a timedelta in the form <int>h, <int>m, <int>s or a combination of the three. Default 10s.

frontend_timeout: str

Time before requests sent to coco time out. This value should depend on how many layers your configuration files have. If a call to a coco endpoint could take longer than this value, because it triggers many layered forward calls you should increase this. A string representing a timedelta in the form <int>h, <int>m, <int>s or a combination of the three. Default 10m.

exclude_from_reset: list

A list of strings that are state paths to be excluded from state resets. Default: None