ElasticDW Docs

Command: cluster admin

The cluster admin command provides an interface to interact with administration tools provided by the data warehouse in a safe and gated way that maintains the stability of the cluster. For example, the Vertica admintools utility. This is necessary, since direct OS level access to managed service nodes is not provided, this is a convenient way to run common (non-distructive commands).

The way the command works is that it provides a set of admin utility commands that can be initiated by the command line utility. You pick the command from a list and it handles running the command on the specified cluster and returns the results.

Usage

edw cluster admin -describe

edw cluster admin \
  -cmd <id_of_command_to_run> \
  -cluster-id <cluster_id> \
  -environment-id <cluster_environment_id>

edw cluster admin \
  -status-id <id_of_running_comand>  \
  -cluster-id <cluster_id> \
  -environment-id <cluster_environment_id>

Required Flags

  • -cmd - (Required if initiating a new command) The ID of one of the available admintools commands to run. This will run for a period of time (see -timeout) or til completion if sooner. If it times out the command will provide a status-id that can be used to look up the status of the running command. Cannot be used with -status-id
  • -status-id - (Required if checking on running command) If the command is long running, this id is returned, and the command can be run again to use this id to check on the status of the remote running command. Cannot be used with -cmd
  • -cluster-id - The cluster ID where commands will be executed.
  • -environment-id - The environment ID that this cluster belongs to.
Note: If the environment has not been configured or set via environment variables, global flags are also required.

Optional Flags

  • -describe - Lists the available admintools commands to run. This list will change over time; for the current list of available commands see below.

Available Remote Commands

  • Get all node status
  • Provides the status of recovery operations
  • Get current state of resource pools on each node
  • Get all database columns
  • Get current K-Safety value
  • Get information about the node subscriptions in the database
  • Get the session subscriptions in the database
  • Indicates how up to date the catalog is on communal storage
  • Get all users in the database
  • Get data usage broken by nodes
  • Get data usage by top 10 tables
  • Get license status

Output

If -describe is used, it will provide a list of commands to pick from:

(0) for: Get all node status
(1) for: Provides the status of recovery operations
(2) for: Get current state of resource pools on each node
...

If -cmd is used, it will provide the ongoing Status til a time out is reached, or the command completes, in which case it will provide the remote commands output. Calling -status-id on an already running command will also provide a similar output.

Status: InProgress
Status: InProgress
Status: Success
Output:
  DB     | Host | State
--------+------+-------
  cli_db | ALL  | UP

Example: Describe the available commands

$ edw cluster admin -describe

(0) for: Get all node status
(1) for: Provides the status of recovery operations
(2) for: Get current state of resource pools on each node
(3) for: Get all database columns
(4) for: Get current K-Safety value
(5) for: Get information about the node subscriptions in the database
(6) for: Get the session subscriptions in the database
(7) for: Indicates how up to date the catalog is on communal storage
(8) for: Get all users in the database
(9) for: Get data usage broken by nodes
(10) for: Get data usage by top 10 tables
(11) for: Get license status

Example: Run one of the available commands - Get all node status

$ edw cluster admin \
    -cmd 0  \
    -cluster-id <cluster_id> \
    -environment-id <cluster_environment_id>

Status: InProgress
Status: InProgress
Status: Success
Output:
  DB     | Host | State
--------+------+-------
  cli_db | ALL  | UP