The project is in a healthy, maintained state
Retry a ShopifyAPI request if if rate-limited or other errors occur. Works with the REST and GraphQL APIs.
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
 Dependencies

Development

~> 5.0
~> 12.0
 Project Readme

Shopify API Retry

CI

Simple Ruby module to retry a ShopifyAPI request if rate-limited or other errors occur. Works with the REST and GraphQL APIs.

Installation

Bundler:

gem "shopify_api_retry"

Gem:

gem install shopify_api_retry

Usage

REST API

By default requests are retried when a Shopify rate limit error (HTTP 429) is returned. The retry happens once after waiting for the seconds given by the HTTP Retry-After header:

require "shopify_api_retry" # requires "shopify_api" for you

ShopifyAPIRetry::REST.retry { customer.update_attribute(:tags, "foo") }
customer = ShopifyAPIRetry::REST.retry { ShopifyAPI::Customer.find(id) }

You can override this:

ShopifyAPIRetry::REST.retry(:wait => 3, :tries => 5) { customer.update_attribute(:tags, "foo")  }

This will try the request 5 times, waiting 3 seconds between each attempt. If a retry fails after the given number of :tries the last error will be raised.

You can also retry requests when other errors occur:

ShopifyAPIRetry::REST.retry "5XX" => { :wait => 10, :tries => 2 } do
  customer.update_attribute(:tags, "foo")
end

This still retries rate limit requests, but also all HTTP 5XX errors.

Classes can be specified too:

ShopifyAPIRetry::REST.retry SocketError => { :wait => 1, :tries => 5 } do
  customer.update_attribute(:tags, "foo")
end

You can also set global defaults.

GraphQL API

By default a retry attempt is made when your GraphQL request is rate-limited. In order to calculate the proper amount of time to wait before retrying you must set the X-GraphQL-Cost-Include-Fields header (maybe this library should do this?):

require "shopify_api_retry" # requires "shopify_api" for you

ShopifyAPI::Base.headers["X-GraphQL-Cost-Include-Fields"] = "true"

Once this is set run your queries and mutations and rate-limited requests will be retried. The retry happens once, calculating the wait time from the API's cost data:

result = ShopifyAPIRetry::GraphQL.retry { ShopifyAPI::GraphQL.client.query(YOUR_QUERY) }
p result.data.whatever

To calculate the retry time the query's return value must be returned by the block.

You can override the retry times:

result = ShopifyAPIRetry::GraphQL.retry(:wait => 4, :tries => 4) do
  ShopifyAPI::GraphQL.client.query(YOUR_QUERY)
end
p result.data.whatever

Like retry attempts made with the REST API you can specify errors to retry but, due to the GraphQL specification, these must not be HTTP status codes and are only relevant to network connection-related errors:

result = ShopifyAPIRetry::GraphQL.retry SocketError => { :wait => 1, :tries => 5 } do
  ShopifyAPI::GraphQL.client.query(YOUR_QUERY)
end

To give wait and try times for specifically for rate limit errors, use the key :graphql:

result = ShopifyAPIRetry::GraphQL.retry :graphql => { :wait => 10, :tries => 5 } do
  ShopifyAPI::GraphQL.client.query(YOUR_QUERY)
end

Note that specifying a :wait via :graphql or without an error key will skip calculating the retry based on the API response's cost data.

Global Defaults

You can configure global defaults to be used by REST and GraphQL calls. For example:

ShopifyAPIRetry.configure do |config|
  config.default_wait  = 2.5
  config.default_tries = 10

  # Use default_* for these
  config.on ["5XX", Net::ReadTimeout]

  config.on SocketError, :tries => 2, :wait => 1
end

See Also

License

Released under the MIT License: www.opensource.org/licenses/MIT


Made by ScreenStaring