Final test: SRE practices: SLOs, incidents, redundancy

Question 12 of 15

One correct answer

With the sick copy 8082, every other answer was `500`, but with the stopped copy there were no failures at all. Why?

systemd warned nginx that the service had stopped.
nginx removed the stopped copy from the upstream in the file right away.
The sick copy answered more slowly, and clients did not wait for the answer.
A refused connection is in the default proxy_next_upstream, while a 500 answer is not.

Back A–D or ↵ Sign in to answer