Tasks failing silently with error 'Instance Unavailable'

[bmurr]

Tasks have been failing silently in my taskqueue all day.
No PermanentTaskFailures or other exceptions are being thrown -- nothing in my logs.
The tasks are just retrying a number of times, with the reason for retries being 'Instance Unavailable'. Then they disappear.


I received an email from AppEngine Support about a month ago about a bug they were planning to fix:

 

Currently, we have a bug in the system where tasks enqueued from a custom domain (e.g. your_domain.com) are always routed to the default module even if a non-default module is specified in the app's dispatch.yaml/xml file. This bug will be fixed by February 22, 2016.

However, this email did not mention the app I'm having issues with and the bug didn't seem relevant to my app.

Were changes made to the TaskQueues today? Is anyone else experiencing similar problems?

[Nick (Cloud Platform Support)]

Does the issue you're experiencing seem like it would be related to the behaviour described in the communication? Is the "instance unavailable" error possibly related to default/non-default modules?

If you experience what you believe to be a platform issue, not the fault of your own code, you should post to the Public Issue Tracker. This forum is meant for more general discussion of the platform and services.