AS216394 · WireShield

Technical Guide

What is anti-DDoS? How it works and how to choose the right solution

Anti-DDoS is a protection system that detects attack traffic, filters it and forwards only clean traffic to the target. Learn how it works, the solution models and the selection criteria.

Anti-DDoS filters attack traffic and forwards only clean traffic to the server

Anti-DDoS is the name for protection systems that detect distributed denial-of-service attacks, filter the attack traffic and forward only legitimate traffic to the target. It is also called "anti-DDoS protection", "DDoS mitigation" or "DDoS scrubbing". The goal is to keep the service available to users even while an attack is under way. This guide answers what anti-DDoS is, which steps it goes through, which solution models exist and what to look for when choosing. You can read about the attack itself first in what is a DDoS attack.

What is anti-DDoS?

An anti-DDoS solution is a layer where traffic is analysed before it reaches the target and the part classified as attack is dropped. Its difference from a network firewall is that it can handle volumetric and protocol attacks in front of your link or at the network edge, at very high packet rates. We explained why a firewall alone is not enough in what is a firewall.

How does anti-DDoS work?

Most solutions work in four steps:

  1. Detection: Traffic is analysed continuously, looking for abnormal behaviour at protocol and vector level, for example a sudden jump in the UDP packet rate toward a target or a pile-up of SYNs from unverified sources.
  2. Diversion: In the remote model, traffic is pulled to a filtering network with BGP. In the inline model, traffic already passes through the appliance.
  3. Filtering: Attack traffic is dropped and legitimate traffic passes. The techniques used in this step are below.
  4. Delivery: Clean traffic returns to the infrastructure over GRE, VXLAN or a cross-connect.

Filtering techniques

In WireShield protection profiles these filters work together:

  • SYN validation: SYNs from unverified sources are answered with a deliberately wrong reply; a real client responds with an RST and is verified. Spoofed sources cannot pass and no connection state is kept.
  • ACK/RST validation and TCP sanity checks: Packets that do not belong to an existing connection or carry invalid flags are removed.
  • Amplification filter: The large response traffic of reflection attacks such as DNS and NTP is filtered.
  • Spoofed source filter: Packets with forged source addresses are dropped.
  • Threat intelligence: For categories such as botnet, reflector, scanner, Tor exit and open proxy you choose Monitor, Rate limit or Drop.
  • Country/ASN policy and payload rules: Custom rules are written by traffic source and content.

You can find how these filters work in detail in rules, lists, GeoIP and threat intelligence in the documentation.

Anti-DDoS solution models

ModelHow it worksWho it suits
Remote (off-site)Traffic is routed over BGP to the filtering network and the clean part returns to youThose who want to keep their current data center and prefixes, and those with high volumetric risk
Inline (network edge)The appliance sits between the uplink and the internal network and filters on siteThose who want traffic to stay inside their network
Web (L7) protectionDNS is pointed to the service; a WAF, rate limiting and challenge pages filter HTTP trafficWebsites, APIs, customer portals

To decide, see inline or remote. The models can be planned together; we decide in the technical discussion which traffic is filtered where. WireShield's three protection models are described on the remote protection, WireShield Edge and Web Protection pages.

Questions to ask when choosing an anti-DDoS solution

  1. How much capacity? Does the upstream capacity of the filtering network exceed the attack volume you expect? WireShield's upstream capacity is above 2 Tbps and the filtering network has four PoPs in Istanbul, Frankfurt, Amsterdam and Sofia.
  2. Which attack types does it cover? Are volumetric, protocol and application-layer attacks handled by the same solution, or separately?
  3. Can you protect your existing infrastructure? Can you use it always-on or on-demand without moving your servers?
  4. What is the false-positive risk? Is there a monitoring mode and fine-tuning so that legitimate traffic is not blocked?
  5. How is visibility? Are attack history, live traffic, source distribution and reports provided? We covered this in reading an attack report.
  6. Integration: Can you connect to your own systems with e-mail, webhook and API?
  7. Onboarding and support: Are BGP, tunnel and certificate setup planned together?

Common mistakes

  • Looking for anti-DDoS after the attack has started.
  • Protecting only the network layer and leaving the web application exposed.
  • Choosing capacity only for average traffic and ignoring peaks and attack moments.
  • Skipping the backup route and the go-live test. The BGP go-live checklist was written for exactly this.

Frequently asked questions

What is the difference between anti-DDoS and a firewall?

A firewall allows or blocks traffic according to rules and fails once the link is full. Anti-DDoS is designed to filter attack traffic at high speed in front of the link or at the edge.

Does anti-DDoS protection slow my site down?

Because traffic passes through a filtering network, a very small delay may be added. That effect cannot be compared with the service being completely unreachable under attack.

Should I use it always-on or on demand?

Always-on protection reacts to an attack immediately; on-demand protection only starts when needed. We decide together in the technical discussion based on your risk profile.

Do I need my own IP addresses for anti-DDoS?

Remote protection works with your own ASN and prefixes over BGP. If you have no ASN, describe your situation and we will evaluate the options together. Inline protection does not require an ASN.

Do websites need anti-DDoS too?

Yes. Application-layer attacks such as an HTTP flood also need a WAF and rate limiting, which Web Protection provides.

Next step

Describe your network, prefixes and web applications and we will suggest the anti-DDoS model and connection method that fits. You can look at the services on the WireShield home page or talk to the technical team.

All posts

Let’s review the right protection model for your network together.

Contact our team to talk through your technical needs.