API World + CloudX + Dev Innovation Summit 2024

Vicky Deliwala, Cisco Systems, Software Engineering Leader

In the modern era of the cloud environment, users needs to think about handling the temporary service interruptions caused by API and micro-services failure.
As the applications interact with remote services and cloud resources, transient faults are more likely to be encountered.
Events such as API’s throttle limits, connection failures, Quota exhaustion, power outage etc. can lead to such unwanted
faults and not so good user experience.

The best practice in such scenarios is to come up with an automated and intelligent retry mechanism which can hide the temporary exceptions from the customers. Challenges faced while designing such strategy are fault identification, tracking
Retry counter, appropriate log instrumentation and custom fallback. This presentation will focus on defining a smart Retry strategy along with specific guidelines to overcome
such challenges gracefully. Through a deep dive into real world scenario, this session will walk-through the design/architecture of RESTful API’s to provision OpenStack VMs along with NoSQL. It will showcase the profound Retry Logic implementation
using the Cron job and scheduler which can significantly improve user experience.

Most important things that people will learn from presentation:

- Transient failures in cloud native apps and how to utilize smart and profound logic to mitigate those challenges.
- Walk through the workflow of a real life scenario to understand the best practices in integrating automated Retry strategy while designing API driven micro-services.