Hetzner includes automated DDoS protection at no extra charge for its customers. The system inspects traffic at the network edge, recognises attack patterns by traffic volume and packet characteristics, and filters common volumetric attacks before they reach your server. It is primarily network and transport layer protection, so firewall rules, rate limiting and application hardening remain your responsibility.
What DDoS Protection Means in the Hetzner Context
Hetzner uses automated security tooling to protect web applications, websites, servers and IT infrastructure from distributed denial-of-service attacks. The protection is built into the network rather than sold as a separate product, and Hetzner states it is available to all customers without raising prices.
The practical distinction that matters most is between network-level filtering and application-level defence. Network-level protection deals with floods of packets and connections: the kind of traffic that tries to saturate bandwidth or exhaust a server's connection table. Application-level attacks look like ordinary requests to your web service and must be filtered by something that understands your application, such as a web server configuration, a web application firewall or a CDN.
How the Filtering Works
Hetzner's system uses Arbor and Juniper hardware as part of its mitigation infrastructure. It recognises attacks by the amount of traffic or the number of packets involved, and then identifies the specific attack type so it can respond to that pattern rather than applying a generic block.
Volume alone is not the trigger. Hetzner gives the example that a UDP flood at 500,000 packets per second is harmless for a server, while the same rate of SYN packets could cause real problems. That is why the system classifies the attack before deciding how to react.
Scrubbing Known Attack Patterns
Traffic scrubbing filters remove commonly known attack traffic. Hetzner says this method is especially successful against DNS reflection, NTP reflection and UDP floods on port 80. Reflection attacks work by abusing third-party services to send large responses to a spoofed target, so filtering them at the edge keeps that junk away from your machine.
Challenge-Response and Dynamic Filtering
In a further layer, Hetzner filters SYN floods, DNS floods and invalid packets using challenge-response authentication and dynamic traffic filtering. Legitimate clients can answer the challenge and continue; automated flood traffic typically cannot. Hetzner says this dynamic method means customer traffic is usually not affected.
What the Protection Covers
The explicitly documented coverage is concentrated at the network and transport layers. Hetzner's own description lists filtering for DNS reflection, NTP reflection, UDP floods on port 80, SYN floods, DNS floods and invalid packets. These are the attacks that can be identified from packet headers and connection behaviour without inspecting application payloads.
Hetzner also states that its automated system recognises almost all attack patterns in advance and blocks the vast majority of attacks. Its detection runs continuously, and its recognition improves as technicians analyse each attack and adjust filters.
What Is Not Covered
Application-layer attacks are the main gap. A slow HTTP request flood, a credential-stuffing run against your login form or a scraper hammering an expensive search endpoint can look like normal traffic at the network edge, and Hetzner's documented filtering does not describe inspecting that layer. You should treat it as your responsibility to defend.
This is also why the included protection is best understood as a baseline. It removes a large class of attacks that would otherwise overwhelm your server's network path. It does not tune itself to your specific application, and it has no view of whether a request is legitimate for your service.
Your Side of the Defence
Hetzner publishes guidance on protecting game servers from DDoS attacks by setting up a basic firewall and rate limiting. The same principles apply more broadly.
- Firewall rules. Allow only the ports and protocols your service actually needs. Every open port is an invitation.
- Rate limiting. Cap requests or connections per source and per endpoint so a single client cannot monopolise capacity.
- Connection limits. Bound concurrent connections so a connection flood cannot exhaust your server's resources.
- Application hardening. Keep services patched, avoid expensive unauthenticated endpoints where possible, and log enough to recognise abuse patterns.
These measures complement the network filtering rather than replacing it. They catch the traffic that looks legitimate at the packet level but is abusive at the application level.
During and After an Attack
Hetzner describes its mitigation as dynamic, so in most cases your legitimate traffic continues to flow while the attack is being filtered. Detection operates continuously, and the filters are adjusted over time as new attacks are analysed.
After an attack, review your own logs and metrics as well as any alerts from your monitoring. Look for connection spikes, unusual request rates, error bursts and changes in traffic sources. Because the protection is automatic, the useful lesson usually comes from what your service saw rather than from anything you need to configure.
When to Add Another Layer
The built-in protection is likely sufficient if your main risk is volumetric flooding: large floods of packets aimed at saturating your bandwidth or exhausting connections. The automatic filtering is designed for exactly that, and it costs nothing extra.
Consider additional protection if your service is exposed to application-layer abuse, if you need request-level rules specific to your application, or if you want tighter control over what reaches your origin. A reverse proxy, a CDN with a web application firewall, or a specialised scrubbing service can sit in front of your Hetzner server and handle the layer that Hetzner's network filtering does not.
One practical caution: Hetzner's acceptable-use policy prohibits attack traffic, and its systems can act on traffic that resembles an attack, so running your own DDoS test can trigger the protection. Coordinate with Hetzner before any such test rather than firing traffic at your own server.
Conclusion
Hetzner's included DDoS protection is a free, automatic network-edge layer that filters large volumetric attacks and common reflection and flood patterns. It is a strong baseline for most workloads. It is not a complete security service: application-layer defence, firewall rules, rate limiting and a layered front end for sensitive services remain your responsibility. Use the built-in protection to handle the flood, and build your own controls for everything that looks like normal traffic.


