DNS failover

Fail away from unhealthy endpoints with Aptranet Cloud DNS health checks, a 1-second minimum TTL and 210+ Anycast DNS servers.

210+Anycast DNS servers
210+Points of presence worldwide
1 secondMinimum TTL
30 msAverage network latency

DNS failover

DNS should stop handing out dead endpoints.

If the origin or region fails and DNS still returns it, every client retries the outage. Failover that waits on a 300-second TTL is not failover.

Health checks monitor endpoints continuously. Unhealthy targets are removed from eligible answers. A one-second minimum TTL lets planned or automatic changes take effect quickly.

Why teams use Aptranet

What DNS failover looks like on this network.

Cloud DNS answers on 210+ Anycast servers, with GeoDNS, health checks and a one-second minimum TTL for change.

Health-aware records

Non-responsive infrastructure is taken out of the answer set.

One-second minimum TTL

Resolvers can pick up a new answer without waiting on a long cache.

DDoS-protected authority

Failover still works if the DNS layer itself is under volumetric attack.

How to set it up

A cutover you can validate before DNS moves.

Create the Cloud DNS configuration, prove it on an Aptranet hostname, then point production DNS when the path looks correct.

  1. 1
    List the endpoints that can take traffic

    Primary and standby application or origin addresses.

  2. 2
    Attach health checks

    Monitor each endpoint and exclude it when the check fails.

  3. 3
    Set a TTL that matches the RTO

    Use a low TTL on records that must move quickly.

  4. 4
    Rehearse a failover

    Fail a check in staging and confirm resolvers stop receiving that answer.

Outcomes

What changes once the hostname is on Aptranet.

  • Dead endpoints can disappear from DNS
  • TTLs can be as low as one second
  • Standby capacity is only used when needed
  • The DNS layer remains reachable under attack

Frequently Asked Questions

No. DNS failover changes the record resolvers receive. BGP failover changes how an address is announced. Many architectures use both.

Some resolvers clamp TTLs. Design for that, and use health-aware answers so even a clamped cache is less likely to point at a dead target.

Yes. Combine health checks with GeoDNS or weighted records so a region can be taken out of the eligible set.

CDN origin groups protect delivery. Cloud DNS failover protects the hostname that clients resolve before they reach a CDN or origin.

Put DNS failover on the Aptranet edge.

Get started in the Management Console, or talk through origin, DNS and cutover with Aptranet.