A server becomes unstable after a maintenance change. What is the strongest initial response?

1. Erase event logs before review
2. Correlate event logs and performance evidence with the recorded change, then use the tested recovery plan
3. Conclude every error is caused by the most recent user sign-in
4. Apply several unrelated changes simultaneously