A DNS change points a service to a new endpoint, but some users continue to reach the old address. What should be considered before assuming Route 53 failed?

1. An updated authoritative record instantly flushes every resolver cache worldwide.
2. The old application must be unreachable because DNS records physically disconnect servers.
3. Changing a security group Name tag forces every client to resolve again.
4. Recursive resolver and client caching can retain the prior answer until its cached TTL expires.