Replicante Core provides a large set of configuration options with reasonable defaults. This supports common use cases where only a handful of options need attention as well as advanced setup where users have the power to fine tune most details.
Some options do not have reasonable defaults so users will have to set them explicitly:
coordinator.backend and coordinator.options.events.stream.backend and events.stream.options.tasks.backend and tasks.options.storage.backend and storage.options.All options are documented in the example configuration file at the root of the repo.
This file shows all options with their defaults and explains their meaning and available settings. As mentioned above, common use cases should be able to ignore most options and focus on the minimal configuration example.
In most cases, details of these options are documented in the features they influence.
Options that don’t directly relate to a single feature are instead documented below.
Replicante Core is a distributed system where components may be owned by different parties. As such we often expect things to happen within time limits but there is no guarantee that they will in fact happen at all.
Timeouts are used to ensure that malfunctioning, slow, or unresponsive elements (agents, processes, dependencies, …) do not permanently or severely impact the entire system.
The timeout related options allow operators to tune the level of sensibility:
Available timeout options:
timeouts.agents_api: timeout applied to all agent HTTP requests.Some of Replicante Core dependencies (DB, coordinator, …) also support timeouts that are documented along the other options for each dependency.