We are currently revamping our website to improve user experience. During this time, some pages or links may not function as expected. We regret any inconvenience and appreciate your patience and continued support.

DDoS Readiness Checklist for Websites, Portals and APIs

A DDoS readiness checklist for IT heads with public sites that earn sales or leads. Eleven items are run for each website, portal and API before any flood, and each has a plain ready when line, an owner and a note on what a gap means. The matching AWS and Azure setting sits beside each item, so a team can see what to look for on either cloud. Nothing is scored, and no prices appear.

A DDoS readiness checklist tells you, before any flood, whether each website, portal and API has filtering in front, alerts that reach people and a team who know what to do. This one has 11 items. Each is either ready or a gap, and each has an owner and the AWS and Azure setting.


Last reviewed 2026-11-16 by Manish Sharma.

Run it once per public service, not once per company, because each service has its own exposure. A portal and an API on the same domain get their own pass, because their risks differ. The pass is DDoS attack preparation in practical form. Also, how Signisys keeps public sites answering under heavy traffic shows where it fits in a wider program. Teams that want to prepare for a DDoS attack in order can still start at the top. DDoS readiness is the sum of the 11 lines below.

The DDoS Readiness Checklist in One Table

Each row is one item, with its ready line, its owner and the matching setting on each cloud.

Eleven items in three groups: what sits in front of each site, watching traffic and reaching people, and testing and review
Eleven items, three groups.
Item Ready when Owner On AWS On Azure
1. Sites listed Every public site and API is on one list IT lead Endpoints listed per account Endpoints listed per subscription
2. Network filtering Each address sits behind it Network lead AWS Shield Azure DDoS Protection
3. Front door Public traffic enters only through it Platform lead Amazon CloudFront Azure Front Door
4. Web rules Block mode on, limits per route Application owner AWS WAF Azure Web Application Firewall
5. Origins closed A direct request gets no answer Network lead Private subnet origins Private connection from the front door
6. Alerts A test alert reached a named person Operations lead Alarms tied to AWS Shield Under-attack alert
7. Provider route Route recorded and used once IT lead AWS Shield Response Team DDoS Rapid Response team
8. Team Names, deputies, call order written IT head AWS support contacts Microsoft support contacts
9. Rehearsal Run in the agreed period, gaps noted IT head AWS testing policy Approved test partners
10. Load test Last result on file Application owner Above-normal load test Scale out, load test
11. Review Dated write-up, changes tracked IT head Per event Per event

DDoS Readiness Checklist: What Sits in Front of Each Site

The first five items decide whether a flood reaches the servers at all.

Ready or a Gap

Every item has two states. Ready means the line is true today. Otherwise the item is a gap with a named owner.

Each Public Site, Portal and API Listed

Ready when one list names every public hostname and API endpoint, with its owner, where it runs and what is exposed. Microsoft’s guidance on a DDoS response strategy asks teams to ask, on an ongoing basis, which new public resources need protection.

Owner: IT lead.

If missing: a site nobody listed stays unfiltered, and the first sign is the outage.

Network-Level Filtering Switched On

AWS Shield covers network and transport layer floods and applies automatically on AWS. On Azure, however, a team turns on Azure DDoS Protection for the virtual network or the public IP address.

Ready when each site’s address sits behind it.

Owner: cloud or network lead.

If missing: a flood fills the link before any rule runs.

Key Takeaway Filtering and a front door decide whether a flood reaches the servers at all.

A Front Door With Caching Ahead of the Servers

Put Amazon CloudFront, or Azure Front Door, in front of each web service. Then turn on caching for pages that allow it. The Microsoft Learn page on Azure Front Door and DDoS says caching shields the backends from large volumes. If an edge service answers DNS, Amazon Route 53 fits that line on AWS.

Ready when public traffic reaches the site only through the front door.

Owner: platform lead.

If missing: traffic goes straight to the servers, with nothing in between to absorb it.

Web Firewall Rules and Request Limits

Use AWS WAF or Azure Web Application Firewall with managed rules and rate limits on login, search, checkout and API calls. A Layer 7 flood is a stream of normal-looking web requests. Network filtering alone does not stop it.

Ready when the rules are on in block mode and each route has a limit.

Owner: application owner.

If missing: a flood of ordinary requests reaches the login page and the checkout.

Origin Servers Kept Off the Open Internet

Servers should answer only the front door, while only the needed ports and protocols stay open. On AWS, the edge service can fetch from origins in private subnets, so the origin need not face the internet. On Azure, however, a private connection runs from the front door to the origin.

Ready when a direct request to the origin address gets no answer.

Owner: network lead.

If missing: an attacker who learns the address skips the front door.

Closed Origins

A front door protects nothing if the origin still answers direct requests. Test it from outside.

DDoS Readiness Checklist Items for Watching Traffic and Reaching People

The next three items decide how fast the right person learns of a flood, and then acts. Also, each one needs a name beside it.

Normal Traffic Known, Alerts Routed to Names

Record normal traffic for each site, and send alerts to named people by more than one channel. On Azure, for example, the “Under DDoS attack or not” alert can reach people by email, SMS, push or voice. On AWS, alarms in AWS’s monitoring service link to what AWS Shield detects. For web request floods, the AWS page on responding to DDoS events says AWS attempts to detect them and notify customers through those alarms where the protection includes it, and by default it does not apply mitigation on its own, to avoid blocking valid users.

To check for an attack, compare live traffic with the recorded normal, and also read the provider’s alert. Also, what a flood looks like from the outside lists what to look for.

Ready when a test alert reached a named person.

Owner: operations lead.

If missing: the first alert is a customer message.

Provider Contacts and Support Route on Record

Write down the support route on AWS and on Azure. Then note the AWS Shield Response Team and the DDoS Rapid Response team, and check whether your setup includes each.

Ready when the route and account details sit in the runbook and one person has used them.

Owner: IT lead.

If missing: during a flood, time goes on finding a phone number and a contract.

A Named Team Ready to Act

Name a response team with contacts for planning and for acting. Choose people who know the protection service and can work with the provider’s support team. Then add deputies and a call order.

One question belongs here and stays open. Who tells customers and staff that a site is slow, and through which channel?

Ready when the team has written down names, deputies and a call order.

Owner: IT head.

If missing: the on-call person is alone, and the customer message waits.

Write the Call Order

A call order on one page saves minutes when a flood starts.

Testing and the Review After in the DDoS Readiness Checklist

The last three items turn a plan into something the team has used.

Rehearsals Kept Inside the Provider’s Test Rules

Simulation exercises, with scale testing, belong in continuity planning. Still, the provider sets the rules. On Azure, simulation runs only with Microsoft’s approved testing partners and only against Azure-hosted public IP addresses in your own subscription that sit under Azure DDoS Protection. On AWS, simulation of network or transport floods follows the AWS testing policy, which requires a testing partner that AWS has approved in advance.

Ready when a rehearsal ran in the agreed period and the team wrote down the gaps.

Owner: IT head.

If missing: the first real test is the first real attack.

Load Tested Above Normal Traffic

Test at the expected traffic and above it, including a sudden crowd, and test again before a major release. Scale out across more than one instance, so one server is not the limit.

Ready when the last test result is on file with the traffic level reached.

Owner: application owner.

If missing: a busy launch day looks like an attack, and nobody knows the limit.

A Review Written After Each Event

After each attack or rehearsal, record what failed, which services suffered, how well the response went and what changes. The write-up feeds the DDoS response plan, so each event improves the next one.

Ready when each event has a dated write-up and the team tracks its changes.

Owner: IT head.

If missing: the team finds the same gap twice.

DDoS Readiness Checklist Gaps for a Website, Portal and API

Count nothing and calculate nothing. List the gaps per service, each with its owner, and then hand the list to the team. A site that earns sales or leads goes first.

Signisys’s working view is to close gaps in front of the site (items 2 to 5) before testing items. The basis is that a rehearsal on an unfiltered site teaches little, as the flood still reaches the servers. A separate article covers the order of work to close each gap.

The three kinds of service differ in where their gaps show up.

DDoS readiness checklist gaps differ by service: caching for a website, login limits for a portal, per-key limits for an API
Gaps show up in different places for each kind of service.

A website leans on caching and the front door. In contrast, a portal leans on the login route and its request limits. An API leans on rate limits per client and per key, and its alerts should watch error rates as well as volume.

Owners and Review Dates for the DDoS Readiness Checklist

The IT lead runs the list, while each item has its own owner, who answers for it. However, a gap without an owner is not yet a gap anyone can close.

Run the list again when a new public site goes live, when a site moves to another provider, and after each event. Microsoft’s guidance asks teams to understand risk on an ongoing basis, so we claim no fixed interval here. Put the date of each run beside the list, so the team sees how old it is.

Key Takeaway A dated list with a named owner on every line is what makes a gap closable.

Where the DDoS Readiness Checklist Stops

The list says what is ready and what is not. The article linked above covers how to close each gap. Choosing between network filtering and a web firewall is set out in a side by side look at the two layers.

Signisys provides cybersecurity services and a network and endpoint security solution, and is also a partner of AWS and Microsoft. The DDoS Readiness Review is a working session. It runs the 11 items on your websites, portals and APIs, then lists the gaps with owners, and findings decide the rest. No charge, with a fixed scope.

Book a consultation
Run the list on your public sites with us

DDoS Readiness Checklist: FAQ

Does a site already behind an edge service still need the other ten items?

Yes. The front door covers item 3, and it also helps with caching and absorbing load. However, origins, alerts, the team and the tests still apply, and the Azure Front Door page itself advises protecting the origin virtual network as well.

Which items change for an API with no web pages?

Item 4 leans on rate limits per client or key, while caching in item 3 may not apply. Item 6 also adds alerts on error rates as well as volume. The rest of the list stays the same.

What should prompt a fresh pass through the list?

A new public site or API, a move to another provider or region, or a large launch or campaign. Any of these changes the exposure, so the earlier answers no longer describe the estate.

Can we run our own flood test against our sites?

Only inside the provider’s rules. On Azure, that means Microsoft’s approved testing partners against Azure-hosted public IP addresses in your own subscription that are under protection, and on AWS it means the AWS testing policy, which requires a testing partner that AWS has approved in advance. We describe no method or tool here.

Newsletter

Security Insights, Delivered.

Practical, vendor-neutral analysis of threats, tools, and architecture decisions.

Book a Consultation

A practice principal will contact you within 1 business day.

Thank you! A Signisys principal will be in touch within 1 business day.