retry

Retry Microservice

📄 Contents

• Contents
  • Description
  • Deployment
  • How does the retry microservice work?
  • Directory description
  • File description
  • Common questions

❕ Description

The Retry service's goal is to reschedule pods that failed

❕ Deployment of the retry service

Before deploying the retry.yaml file, please configure the environment variables to the correct values used by the queue microservice.

env:
- name: MQ_HOST
  value: "sched-rabbitmq-0.sched-rabbitmq.custom-scheduler.svc.cluster.local"
- name: MQ_PORT
  value: "5672"
- name: MQ_USER
  value: "guest"
- name: MQ_PASS
  value: "guest"
- name: RECEIVE_QUEUE
  value: "epsilon.backoff"

❕ Retry algorithm

[STEP 1]
The retry service monitors the queue for new pods that failed.

[STEP 2]
When a pod that failed is recevied, the retry service will generate a backoff timer and wait for the backoff timer to pass

[STEP 3]
Once the backoff duration had past, the retry service will send the failed pod back to its respective scheduling queue

❕ Directory and File Description

❕ Common questions

How to change the retry algorithm?

The function WaitAndSend() in line 176 of main.go, contains the implementation of the retry algorithm. By modifiying this function the retry algorithm can be modified.