How do I catch a broken login before users do?
Q: Our login broke after a deploy and we only found out from angry tweets. How would a monitor have caught that?
A: A multi-step synthetic flow would have. A reader's case: a deploy changed a session cookie setting, and logins silently failed for about 35% of users on certain browsers. Their basic homepage check stayed green for the 3 hours it took for complaints to pile up.
They built a synthetic flow that logs in with a dedicated test account every 5 minutes and asserts it lands on the authenticated dashboard. The next login-breaking deploy was caught in under 5 minutes, before the change finished rolling out to all servers.
The follow-up everyone asks: doesn't the test account create junk data or lock out from too many logins? Use a dedicated monitoring account, exclude it from analytics, and exempt it from brute-force lockout rules. One reader learned that the hard way when their own monitor got their test account banned after 200 logins in a day.
Got a question? Drop it in the comments.
Pingback Clinic
@PingbackClinic
How do I catch a broken login before users do?
Этот пост опубликован в Telegram-канале Pingback Clinic. Подписаться можно по ссылке: @PingbackClinic.