Friday afternoon (New York time) I started seeing dramatic increase in request latency of my Python 2.7 GAE app in production. It looks like it's related to an increase in loading latency, which affects auto-scaling. I made no recent changes to my app, nor have traffic usage patterns changed. It seems to be something that GCP changes in their infrastructure and/or runtime.
I have an open Priority 1 ticket with GCP Support, but they're not divulging that they made a change that caused this. I'm wondering if anyone else in the djangae user group has also experienced this issue?