WEBVTT

00:00.020 --> 00:01.190
Welcome back everybody.

00:01.190 --> 00:07.430
So after I stopped flooding the great submission portal with requests, the horizontal pod autoscaler

00:07.430 --> 00:09.470
brought it back to one pod.

00:09.470 --> 00:14.330
And that makes sense, because you'll remember that CPU usage is a very volatile metric.

00:14.330 --> 00:20.900
It goes up with increasing workload, and when we stopped all of those requests, the CPU went down

00:20.900 --> 00:23.210
because there was no more workload.

00:23.210 --> 00:30.950
As the CPU went down back to healthy levels, the horizontal pod autoscaler scaled it back down to a

00:30.950 --> 00:32.480
single pod replica.

00:32.480 --> 00:39.410
So now we're only at 4% average CPU utilization, which is way below the specified target.

00:39.410 --> 00:41.510
Okay, and that's pretty much it.