Environment Settings Reference¶

This document lists all settings that can be set per environment. These changes are made through the API, the web-console or the CLI tool.

The supported settings are:

auto_deploy¶

Type:: bool
Default:: True

When this boolean is set to true, the orchestrator will automatically release a new version that was compiled by the orchestrator itself.

auto_full_compile¶

Type:: str
Default:: ‘’

Periodically run a full compile following a cron-like time-to-run specification interpreted in UTC with format [sec] min hour dom month dow [year] (If only 6 values are provided, they are interpreted as min hour dom month dow year). A compile will be requested at the scheduled time. The actual compilation may have to wait in the compile queue for some time, depending on the size of the queue and the RECOMPILE_BACKOFF environment setting. This setting has no effect when server_compile is disabled.

autostart_agent_deploy_interval¶

Type:: str
Default:: 600

Set the frequency of deploy runs (i.e. only trigger a deploy for resources that are not compliant as far as we know). When specified as an integer, this will set the wait time (in seconds) before attempting to redeploy a resource after an unsuccessful deployment, on a per-resource basis. When specified as a cron-like expression, a global deploy (i.e. for all resources that have a known divergence with their desired state) will be run following a cron-like time-to-run specification, interpreted in UTC. The expected format is [sec] min hour dom month dow [year] (If only 6 values are provided, they are interpreted as min hour dom month dow year). A deploy will be requested at the scheduled time. Set this to 0 to disable the scheduled deploy runs. When specified as an integer, it must be smaller than the repair interval.

autostart_agent_repair_interval¶

Type:: str
Default:: 86400

Set the frequency of repair runs (i.e. trigger a deploy regardless of the assumed state of the resource(s)). When specified as an integer, this will set the wait time (in seconds) before re-scheduling a resource for deployment after the previous deployment has ended, regardless of success or failure, on a per-resource basis. When specified as a cron-like expression, a global repair (i.e. a full deploy for all resources, regardless of their assumed desired state and regardless of their actual state) will be run following a cron-like time-to-run specification, interpreted in UTC. The expected format is [sec] min hour dom month dow [year] ( If only 6 values are provided, they are interpreted as min hour dom month dow year). A repair will be requested at the scheduled time. Setting this to 0 to disable the scheduled repair runs. When specified as an integer, it must be larger than the deploy interval.

autostart_on_start¶

Type:: bool
Default:: True

Automatically start agents when the server starts instead of only just in time.

available_versions_to_keep¶

Type:: int
Default:: 100

The number of versions to keep stored in the database, excluding the latest released version.

environment_metrics_retention¶

Type:: int
Default:: 336

The number of hours that environment metrics have to be retained before they are cleaned up. Default=336 hours (2 weeks). Set to 0 to disable automatic cleanups.

notification_retention¶

Type:: int
Default:: 365

The number of days to retain notifications for

protected_environment¶

Type:: bool
Default:: False

When set to true, this environment cannot be cleared or deleted.

recompile_backoff¶

Type:: positive_float
Default:: 0.1

The number of seconds to wait before the server may attempt to do a new recompile. Recompiles are triggered after facts updates for example.

reset_deploy_progress_on_start¶

Type:: bool
Default:: False

By default the orchestrator picks up the deployment process where it was when it restarted (or halted). When this option is enabled, the orchestrator restarts the deployment based on the last known deployment state. It is recommended to leave this disabled because in most cases it is faster (because we can skip some redundant work) and it has more accurate state and progress reporting (because we retain more state to reason on). Enable this in case there are issues with restoring the deployment state at restart.

resource_action_logs_retention¶

Type:: int
Default:: 7

The number of days to retain resource-action logs

server_compile¶

Type:: bool
Default:: True

Allow the server to compile the configuration model.