test: randomize the iteration interval

tests/load/README.md

@@ -0,0 +1,16 @@
+The tests use [k6](https://k6.io/) to perform load testing.
+
+# To run the test
+
+Set the following environment variables to point to the deployment.
+
+```
+export WEBAPP_URI=<webapp_uri>
+export SEARCH_API_URI=<search_api_uri>
+```
+
+Once set, you can now run load tests using the following command:
+
+```
+npm run test:load
+```

tests/load/chat.js

@@ -5,8 +5,14 @@ import { group, sleep } from 'k6';
 const chatStreamLatency = new Trend('chat_stream_duration');
 const chatNoStreamLatency = new Trend('chat_nostream_duration');
 
+function between(min, max) {
+  min = Math.ceil(min);
+  max = Math.floor(max);
+  return Math.floor(Math.random() * (max - min) + min); // The maximum is exclusive and the minimum is inclusive
+}
+
 function choose(list) {
-  return list[Math.floor(Math.random() * list.length)];
+  return list[between(0, list.length)];
 }
 
 export function chat(baseUrl, stream = true) {
@@ -55,6 +61,6 @@ export function chat(baseUrl, stream = true) {
     const latencyMetric = stream ? chatStreamLatency : chatNoStreamLatency;
     latencyMetric.add(response.timings.duration, { type: 'API' });
 
-    sleep(1);
+    sleep(between(5, 20));
   });
 }

tests/load/config.js

@@ -1,27 +1,16 @@
 export const thresholdsSettings = {
   'http_req_failed{type:API}': [{ threshold: 'rate<0.01' }], // less than 1% failed requests
   'http_req_failed{type:content}': [{ threshold: 'rate<0.01' }], // less than 1% failed requests
-  'http_req_duration{type:API}': ['p(95)<30000'], // 95% of the API requests must complete below 30s
-  'http_req_duration{type:content}': ['p(99)<500'], // 99% of the content requests must complete below 500ms
+  'http_req_duration{type:API}': ['p(90)<40000'], // 90% of the API requests must complete below 40s
+  'http_req_duration{type:content}': ['p(99)<200'], // 99% of the content requests must complete below 200ms
 };
 
+// 5.00 iterations/s for 1m0s (maxVUs: 50-100, gracefulStop: 30s)
 export const standardWorkload = {
-  executor: 'per-vu-iterations',
-  vus: 50,
-  iterations: 50,
-  maxDuration: '50s',
-};
-
-export const heavyWorkload = {
-  executor: 'ramping-vus',
-  stages: [
-    { duration: '10s', target: 20 },
-    { duration: '50s', target: 20 },
-    { duration: '50s', target: 40 },
-    { duration: '50s', target: 60 },
-    { duration: '50s', target: 80 },
-    { duration: '50s', target: 100 },
-    { duration: '50s', target: 120 },
-    { duration: '50s', target: 140 },
-  ],
+  executor: 'constant-arrival-rate',
+  rate: 5,
+  timeUnit: '1s',
+  duration: '1m',
+  preAllocatedVUs: 50,
+  maxVUs: 100,
 };