Network Operations

If you are under a DDoS attack, do these in order: confirm the attack, identify its scope and type, inform your provider and your protection service, move filtering in front of your link, watch your other systems for secondary attacks and review the incident afterwards. The most important point is to be prepared in advance instead of looking for a solution after the attack has started. This guide summarises the steps to follow during an attack and the preparation to do beforehand.
1. Confirm the attack
Slowness does not always mean DDoS. The US cybersecurity agency CISA notes that symptoms such as slow network performance and unreachable sites can also come from maintenance or technical problems, so they should be confirmed before assuming an attack. Check:
- Has there been a recent change, update or maintenance?
- Is there a sudden, unexplained jump in the traffic graph?
- Is the problem limited to one service or port?
- Do source addresses come from many different countries or networks?
We explained which data matters in the UDP flood monitoring guide.
2. Identify the type and scope
Collect this information to decide:
- Volume: Bits per second (bps) and packets per second (pps). They tell different stories; we explained this in reading an attack report.
- Protocol and port: UDP, TCP SYN, ICMP, HTTP? Reflection traffic usually arrives from service ports such as 53 or 123.
- Target: A single IP, a prefix, a website?
- Source distribution: A narrow pool of networks, or from everywhere?
See types of DDoS attacks to tell them apart.
3. Notify the right parties
- Your network administrator and team: Make it clear who does what.
- Your internet service provider: CISA notes that the network administrator and the provider can confirm whether your network is the target. If the problem is on the provider's side or they are a target too, you may be an indirect victim.
- Your DDoS protection provider: Contact them immediately if you have one. If you are protected by WireShield, reach the technical team about an attack.
4. Move filtering in front of your link
If the link is full, nothing on the server helps. The only way to stop the attack is to filter traffic before it reaches your link:
- Remote protection: Your prefixes are routed over BGP to the WireShield filtering network and clean traffic returns over GRE, VXLAN or a cross-connect. You can use protection always-on or only when needed (on-demand). We described the process in how remote DDoS protection works.
- Inline protection: For attacks that do not fill your link, WireShield Edge filters traffic at your network edge.
- Web applications: For HTTP-level attacks, Web Protection applies rate limiting, a WAF and challenge pages.
5. What not to do
- Block IPs one by one. Sources are many and often spoofed; this wastes time. We explained why in the UDP flood post.
- Keep rebooting the server. It may bring the service up briefly while the attack continues, but it erases logs and clues.
- Respond to ransom demands. Paying does not guarantee the attack stops and can make you a repeat target.
- Watch only one system. CISA notes that attackers can use a denial of service as a distraction for secondary attacks on other services. Monitor your other systems too.
6. Manage communication
Give customers and your team a short, honest status message: what happened, which services are affected and when the next update will be. Do not speculate about who is behind it.
7. Collect evidence
A record of the attack is valuable for later improvement and for any legal process:
- Traffic graphs, timeline and affected services.
- Source distribution, protocol and port details.
- Attack history and report from the portal.
In the WireShield Customer Portal, each attack shows live traffic, vectors, sources, targets and applied mitigations on one screen, and a PDF report can be downloaded.
8. Run a post-incident review
- How long did it take to notice the attack and how long to stop it?
- Which services were affected and what did it cost?
- Why did protection fall short, or which rule worked?
- What changes do rules, thresholds and the communication plan need?
Preparation checklist before an attack
- [ ] ASN, prefix list and LOA/RPKI records are ready.
- [ ] You have decided which traffic is filtered where. See inline or remote to choose.
- [ ] BGP sessions have been tested. The go-live checklist helps.
- [ ] Monitoring and notification channels (e-mail, webhook) are defined.
- [ ] Roles, responsibilities and contacts are written down.
- [ ] Provider contact details and the escalation path are known.
Frequently asked questions
How long does a DDoS attack last?
From a few minutes to days. The exact duration depends on the attacker's goal and how effective the defence is; some attacks return in waves.
Can I get protection after the attack has started?
It is possible, but the service may remain affected during setup and routing. That is why preparation and testing in advance are so important.
Can I find the source of a DDoS attack?
You can see the source distribution, but finding the real attacker is very hard with spoofed and botnet-driven attacks. A narrow ASN pool can be useful for a complaint.
I received a ransom e-mail. What should I do?
Do not pay. Save the e-mail and forward it to your protection provider, and assess together whether the threat is backed by a real attack.
Does it make sense to shut down my server?
Usually not. The service is already unreachable, and shutting down makes collecting data harder. First focus on filtering traffic in front of your link.
Sources
Next step
If you are under attack, or want to prepare in advance, describe your situation: write to us about an attack or plan with the technical team.