SemaphoreCI is currently broken

[cdrage]

SemaphoreCI hasn't been passing lately and it's due to our tests.

Our examples we use for ConfigMap, HealthCheck, etc are using the Wordpress containers which are too resource-intensive for the CI platform, resulting in containers being evicted.

When the containers are being evicted, kube-dns and kube-system are sometimes removed, resulting in negatives in our testing suite.

Hence the reason why sometimes our testing suite passes (the containers just so happen to be up) and other times it fails.

After trouble-shooting multiple things (running in PARALLEL=1, only doing one test at a time) as well as actually logging in and running the testing suite in an SSH console, it's due to the VPS (which is 4GB of RAM I might add) evicting the containers.

So a few solutions to fix this

• Remove Wordpress container as it takes way too long to initialize / first start-up.

• Switch to a less-resource intensive and more standardized test. For example, using the https://github.com/kubernetes/kubernetes/tree/master/examples/guestbook-go guestbook-go example which using go binary + redis clusters

We are currently blocked on #520 and #511 as the direct result of updating our guestbook-go example will result in our tests passing due to them having less resource-constrained containers.

[cdrage]

Fixed! Closing this. Tests are down to 2 minutes! (from 40+ minutes before) 💯 💯 See: https://semaphoreci.com/cdrage/kedge