View Source Timeouts
Timeouts in Ash work a bit differently than other tools. The following considerations must be taken into account:
- If you run a resource action in a transaction, then the timeout applies to the entire transaction.
- If the resource action you are running, and any of its
touches_resourcesis already in a transaction then the timeout is ignored, as the outer transaction is handling the timeout.
- If the resource is not in a transaction, and supports async execution (ash_postgres does), then everything is run in a task and awaited with the provided timeout.
- If the data layer of the resource does not support timeouts, or async execution then timeouts are ignored.
- As of the writing of this guide, none of the API extensions support specifying a timeout. If/when they do, they will run the action they are meant to run in a
Ways to Specify Timeouts
You have a few options.
You can specify a timeout when you call an action. This takes the highest precedence.
MyApi.read!(query, timeout: :timer.seconds(30))
You can specify one using
Ash.Query.timeout/2. This can be useful if you want to conditionally set a timeout based on the details of the request. For example, you might do something like this:
# in your resource defmodule MyApp.SetReportTimeout do use Ash.Resource.Preparation def prepare(query, _, _) do if Ash.Query.get_argument(query, :full_report) do Ash.Query.timeout(query, :timer.minutes(3)) else Ash.Query.timeout(query, :timer.minutes(1)) end end end actions do read :report_items do argument :full_report, :boolean, default: false prepare MyApp.SetReportTimeout end end
And you can specify a default timeout on the Api module that you call your resources with. Overriding an api with a default timeout requires providing a timeout of
:infinity in one of the other methods.
execution do timeout :timer.seconds(30) # this is the default end
Keep in mind, you can't specify timeouts in a before_action or after_action hook, because at that point you are already "within" the code that should have a timeout applied.