theckang
commented
3 years ago For alerts, what would you like to see in the 'end state' that we don't already show? We do simulate a test to make sure the alerts fire, with these sample screenshots to check:
https://raw.githubusercontent.com/RedHatGov/sre-workshop-dashboard/main/workshop/content/images/grafana-alert-test-navigation.png https://raw.githubusercontent.com/RedHatGov/sre-workshop-dashboard/main/workshop/content/images/grafana-alert-test-alerting.png
For example, in Lab 2.1 the lab instructions explain how to create the alert for SLO1, and explains how to create the alert for SLO2 (in the "Click here if you need help" section). At the same time, the "end state" for the alert is not shown , which makes it impossible to correct what I've created as a workshop user, if I did something wrong (and didn't get the results I expected)
Similarly, many of the other labs do provide information about steps to take in order to troubleshoot the expected issues, which is great in order to allow for some exploration and skill building. At the same time, the "Click here for a solution" sections do not clearly indicate the successful end state of the system when the problems are solved. For example, lab 4.2 indicates that the app-ui-autoscale.yml resource didn't specify a replicas field (as the root cause of the issue). However, the "solution" side of the lab never shows the steps to correct that - e.g. a sample of the "corrected" yaml and a command to apply that change. Similarly, lab 5.2 indicates that the solution is to "increase the timeoutSeconds" value, but doesn't show the concrete steps to fix the issue.
Along similar lines, since the ".2" labs are solutions, it might just make sense to leave the solution outside of the "Click here for the solution" section. The labs already ask the user to not proceed to the next lab until they try to fix things on their setup