Retry logic documentation using ElkAPI and OpenAI's official API

Introduction

This document aims to help you understand how to implement automatic request switching and retry logic when using ElkAPI and the OpenAI official API. Requests are prioritized through the ElkAPI. If an error response is received, the request is automatically switched to the OpenAI official API and retried.

Retry Logic Overview

This retry logic consists of two main steps:

Prioritize ElkAPI: First, the API request is attempted using the ElkAPI. If the request succeeds, a response is returned. If the request fails (for example, due to an HTTP error), the second step is performed.

Switch to OpenAI Official API: If a ElkAPI request fails, the program automatically switches to the OpenAI official API for the same request. If the request succeeds, a response is returned. If the request fails, an error message is returned to the user.

Implementation Details

Environment Setup

Before beginning implementation, please ensure that you have installed the necessary Python libraries. This example uses the OpenAI Python client library, so you'll need to make sure it's installed:

Implementation Steps

Here are the specific implementation steps:

Initialize the API Client:

Create a ElkAPI client instance, setting its base_url and api_key.

Create an OpenAI official API client instance, only setting the api_key.

Implement the Request Logic:

Attempt to make a request using the ElkAPI client.

Catch any API errors (such as 400 or 500 errors). These errors are typically caused by a failed request or an error code in the response.

If an error occurs, switch to the OpenAI official API client and make the same request.

For each request method, log and output the error message (if any).

Return a Response or Error Message:

If any request method succeeds, return the request result message.

If both request methods fail, a string containing the error message is returned.

Sample code

The following is a Python sample code that implements the retry logic:

Notes

Error handling: In a production environment, expand the error log to better handle and log errors.

Performance monitoring: Frequently switching to the OpenAI official API may indicate a problem with the ElkAPI, which requires further investigation and optimization.

The implementation of this retry logic is intended to improve the robustness of the system and the request success rate, ensuring that the system can still provide services normally when problems occur at specific API nodes.

Retry logic documentation using ElkAPI and OpenAI's official API

Introduction#

Retry Logic Overview#

Implementation Details#

Environment Setup#

Implementation Steps#

Sample code#