Pivotal Knowledge Base


Update of PCF RabbitMQ Causes PCF Push Notifications Downtime


  • Pivotal Cloud Foundry (PCF) 1.11, 1.12, and 2.0
  • PCF Push Notifications 1.9.2 and below
  • PCF RabbitMQ tile from 1.9.5 to 1.9.6


After upgrading the PCF RabbitMQ tile from 1.9.5 to 1.9.6 with the Push Push Notifications Service errand disabled, the Push Notifications service stop functioning.

In the push-API application logs, you will find something similar to

OUT [ERROR] ..... io.pivotal.push.service.queue.impl.RabbitMQAbstractQueue - [TENANT ...] - 
Failed to enqueue: attempted #1 with exception {}


The upgrade of the PCF RabbitMQ tile will cause the existing Push Notifications service to lose connectivity to Rabbit. Prior to PCF Push Notifications 1.9.3, once this connectivity is lost, the Push Notifications service will not automatically try to reconnect without being restarted.


There are 3 distinct ways to resolve or avoid this problem:

Restart App

Manually restart the "push-API" application via the CF CLI.

Upgrade Push Notifications

Upgrade to a version of PCF Push Notifications greater than 1.9.2.


Have the Push Push Notifications Service errand enabled when you are upgrading PCF RabbitMQ?


Powered by Zendesk