PoC of infra for running ad-hoc scale tests

adhoc/README.md

@@ -0,0 +1,32 @@
+## Ad-hoc scale tests
+
+This directory contains a PoC framework allowing K8s community members to run 
+ad-hoc e2e 100 node scale tests. 
+
+## Prerequisites
+
+**Before using this framework, please contact mm4tt@.** 
+
+[TODO(mm44tt)]: <> (Create a mailing group with owners of this framework.)
+
+## How to run an ad-hoc scale test
+
+1. Implement your test in the `run-e2e-test.sh`. See comments in the file for
+   instructions on how to connect to a cluster or persist the test results.
+2. Send a PR with your changes.
+3. In the PR trigger the test by posting the following comment: 
+   ```
+   /test pull-perf-tests-100-adhoc
+   ```
+4. You can trigger the test multiple times, but only one instance of a test will
+   run at a given time. 
+5. If you need to change your test, update the PR with your changes and trigger
+   the test again.
+6. If you need to change the way cluster is created (e.g. 
+   kube-controller-manager qps settings) send a PR to modify the test definition
+   in https://github.com/kubernetes/test-infra/blob/master/config/jobs/kubernetes/sig-scalability/sig-scalability-adhoc.yaml
+
+## Contact
+
+In case of any questions, reach out to mm4tt@.
+

adhoc/run-e2e-test.sh

@@ -0,0 +1,35 @@
+#!/bin/bash
+
+# Copyright 2020 The Kubernetes Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+set -o errexit
+set -o nounset
+set -o pipefail
+
+# The KUBECONFIG env variable will point to a valid kubeconfig giving you an
+# admin acess to a fresh 100 node cluster. You don't have to tear the cluster
+# down, it will happen automatically after the test.
+export KUBECONFIG="${KUBECONFIG:-${HOME}/.kube/config}"
+
+# If you need your test to persist any output files (e.g. test results) dump
+# them into the /workspace/_artifacts. Content of this directory will get
+# automatically uploaded to gcs after the test completion (successful or not).
+export ARTIFACTS_DIR=/workspace/_artifacts
+
+
+# Implement ad-hoc test here, e.g. install extra addons via kubectl then run
+# clusterloader2 with your custom config.
+
+exit 0