I installed the Kong Kubernetes Controller v1.1.0 with Kong v2.2 on my cluster, running with DBLess mode.
My cluster has about 450K RPS on our core services, and Kong can generally handle that, but recently, I’ve noticed that about 1% of the calls keep getting “no Route matched with those values” for a few minutes, then it promptly returns and acts normally. This tends to happen at our highest scale times, and it goes in and out, occurring consistently for several minutes, then working perfectly, then occurring again after some time, continuing like this for a while.
The Kong Ingress Controller deployment was followed step by step from the Kong for GKE tutorial on the Kong website, and my ingresses are not changing, they haven’t changed in a while.
Anybody have any ideas?