With DNS-based failover between two regions, why can users keep hitting the dead region for a while after failover triggers?
DNS changes propagate only as fast as caches expire, so a 300-second TTL means up to five minutes of misdirected traffic — and some resolvers and applications cache longer than they should. Low TTLs on failover records and client-side retries mitigate this.
🌊 This is a free sample — 1015+ more questions on DevOps Ocean
Sign in free to practice the full question bank with daily challenges, XP, streaks, duels and a global leaderboard.