(Optional) Spike multiple endpoints

Real applications expose many routes; burst capacity is rarely identical across them. A read-heavy cacheable endpoint may handle a spike while a write path or dependency-heavy API degrades first. This milestone extends your spike profile to hit two QuickPizza routes in the same test so you can compare behavior under the same VU ramp.

To keep the per-route metrics readable, this script uses two k6 features:

• Groups (group()) wrap related requests under a label so you can see combined timing for a logical action, such as “home” or “pizza API”.
• Tags (tags: { name: '...' }) attach a label to individual requests so k6 breaks metrics out per route instead of per raw URL, which also keeps metric cardinality low.

To spike multiple endpoints, complete the following steps:

1. Save a copy of your spike.js file as spike-multi.js so you can compare with your single-endpoint run later.

2. Replace the default function with a version that requests both the home page and the pizza API, using group and tags so Grafana Cloud k6 can break metrics out per route without blowing up metric cardinality:

   JavaScript Copy

   import http from 'k6/http';
   import { check, sleep, group } from 'k6';
   
   export default function () {
     group('home', function () {
       const res = http.get('https://quickpizza.grafana.com/', {
         tags: { name: 'GET_home' },
       });
       check(res, {
         'home status 200': (r) => r.status === 200,
       });
     });
   
     group('pizza API', function () {
       const payload = JSON.stringify({
         mustBeVegetarian: false,
         excludedIngredients: [],
         excludedTools: [],
         maxCaloriesPerSlice: 500,
         minNumberOfToppings: 2,
         maxNumberOfToppings: 5,
       });
       const res = http.post(
         'https://quickpizza.grafana.com/api/pizza',
         payload,
         {
           headers: {
             'Content-Type': 'application/json',
             Authorization: 'Token abcdef0123456789',
           },
           tags: { name: 'POST_api_pizza' },
         }
       );
       check(res, {
         'pizza status 200': (r) => r.status === 200,
       });
     });
   
     sleep(1);
   }

   Keep the same export const options block (stages and thresholds) as in your original spike.js.

3. Run the test locally, then send results to Grafana Cloud k6 the same way as in the Visualize the spike in Grafana Cloud k6 milestone:

   Bash Copy

   k6 cloud run --local-execution spike-multi.js

   As in the earlier milestones, drop --local-execution (k6 cloud run spike-multi.js) to generate the load on cloud load generators, which is the better choice for higher VU counts.

4. In the Grafana Cloud k6 result, compare response time and errors per tag or group during the spike phase. The group labels and name tags you set are what let Grafana Cloud k6 split the metrics per route so you can compare them side by side.

   Note which route showed higher latency, errors, or slower recovery. Prioritize that route when you tune pools, timeouts, or rate limits for production.

In the next optional milestone, you add authentication headers driven by environment variables. If you prefer to move on to Turn spike results into action, open that milestone in the sidebar.