
DDoS Stress Test Report: What It Must Contain
A traffic graph is not a DDoS stress test report. The sections a useful report needs, the evidence a reviewer can re-check, and the red flags to reject.
The short answer
A useful DDoS stress test report shows, per attack vector, what was sent, what reached your origin, what the mitigation layer absorbed, how long detection and mitigation took, and what legitimate users felt. It lists ranked fixes and what was not tested. A peak-traffic graph or a pass verdict alone cannot be checked, repeated or defended.
Who this is for
This is for a CTO, head of infrastructure or security lead who is about to buy an authorised DDoS stress test, or who has already received a report and is not sure what it proves. It is also for the person who must show that report to an auditor, a board or a partner. Before the report there is the scope: scoping DDoS stress testing for an online casino covers what to put in scope, and choosing a stress testing provider separates a sanctioned test from a booter. This article is about the last step, the document you are left holding. If you want the same question for the vulnerability side, what a real pentest report looks like does it for penetration testing.
Why a traffic graph is not a report
Most disappointing stress test deliverables fail in the same few ways.
- They describe the attacker, not your defences. A page about how many requests were generated tells you what the provider's tooling can do. It says nothing about which of your controls reacted.
- They collapse everything into one verdict. "Passed" hides which vectors were tested, which were skipped and what a user would have seen.
- They cannot be repeated. Without the timeline, the traffic profile and the configuration in force, nobody can rerun the test after a fix and compare.
- They are written from the provider's view only. The traffic generator knows what it sent. Only your own edge, origin and mitigation logs know what arrived and what was blocked. A report that does not reconcile the two is half a report.
The test itself is only as valuable as the evidence it leaves behind. That is why the report should be specified in the quote, before any traffic is sent.

The report anatomy matrix
Use this as a checklist when you write the quote or review a draft. It contains no thresholds, because the right figures come from your own baseline and your providers' policies.
| Section | What it must show | Red flag if missing | Who relies on it |
|---|---|---|---|
| Scope and authorisation | The signed scope, environment, window, traffic ceiling, rules of engagement and the named person who could stop the test | Nothing ties the test to a written approval | Auditor, legal |
| Timeline | Timestamped start, ramp steps, holds, any pause or stop, and the end, in one time zone | Findings cannot be matched to your own logs | Your engineers |
| Vectors tested | Each attack type and layer (volumetric, protocol, application), with the target it was aimed at | A single "DDoS" line covering everything | Security lead |
| Per-vector behaviour | What reached the origin, what the mitigation layer absorbed, which rule or feature fired | Only an aggregate graph | Infrastructure team |
| Detection and response time | Time from first hostile traffic to alert, and from alert to effective mitigation, with who acted | No clocks recorded, only "mitigated" | On-call owner |
| Legitimate-traffic impact | What test users or synthetic checks experienced during mitigation: errors, latency, blocked sessions | Availability asserted without user-path checks | Product owner |
| Dependencies and side effects | Impact on shared services, partner callbacks, provider notices received | No mention of anything outside the target | System owner |
| Findings and fixes | Ranked list, each with the evidence, a proposed change and an owner | Generic advice that could apply to any site | Engineering |
| Not tested | Vectors, paths and conditions deliberately left out, and why | The report implies full coverage | Everyone |
| Retest plan | What would be rerun after fixes and how a pass would be judged | No route from finding to proof of fix | Auditor, board |
The two rows that separate a real report from a dashboard export are per-vector behaviour and "not tested". One shows what your defences did; the other stops a reader from over-trusting the result.
Per-vector findings: the part you will actually use
For each vector, a reader should be able to answer the same short list of questions without opening an appendix.
- What was sent. Type, target, source addresses, start and end, and the profile (a steady rate, a ramp, a pulse). The aim is repeatability, not bragging rights.
- What arrived. What your edge, load balancer and origin recorded, taken from your own telemetry, not from the generator.
- What stopped it. The mitigation feature, rule or manual action that engaged, and whether it was automatic or needed a person.
- How long it took. The gap between first hostile traffic and a mitigation that actually held. Record who noticed first: your tooling, the provider or a customer.
- What it cost legitimate users. Challenges, blocked sessions, failed requests and slow pages seen by the synthetic users or test accounts you agreed to run during the test.
- What remained exposed. The origin address, a direct-to-IP path or a protocol the mitigation layer did not cover.
If a vector section cannot answer these in a few lines each, the report is describing activity, not results. For a layer 7 example of why per-vector detail matters, see the anatomy in a modern L7 DDoS.
Evidence a reviewer can re-check
A report is only as strong as what stands behind it. Ask for these to be named, even if the raw files are held separately.
- Timestamps in a single time zone, so that your own logs can be lined up against the generator's.
- Source address list or generator identifiers, so that traffic in your logs can be attributed to the test and not mistaken for a real incident.
- Your provider's mitigation events for the window, ideally exported by you from your own account rather than screenshotted by someone else.
- The configuration in force, such as which mitigation rules, rate limits and challenge settings were enabled, because a fix is a change in that configuration.
- The test accounts or synthetic checks used, so that user-impact claims map to something that ran.
For a regulated firm this evidence also has to stand up to a reviewer. The DDoS stress testing sign-off for a digital bank shows how scoped evidence is arranged for a supervisor; the same discipline helps any operator who must answer an auditor.
Findings and fixes: what good looks like
A ranked list of fixes is the point of the exercise. A good finding has four parts: what was observed, why it matters to a legitimate user or to the business, the change that would address it, and who owns the change. Prefer findings such as "the checkout path was challenged for authenticated users once the rule fired, so loosen it for signed-in sessions", over generic advice such as "enable a WAF".
Be sceptical of a list that reads the same for every client. It usually means it was written from a template and not from this test. A useful list is short, ordered by effect on real users, and tied to specific evidence in the timeline.
Red flags when you review a draft
- A report headed by the largest number the provider could send.
- No mention of which of your own logs were used.
- Findings that do not name a vector, a path or a rule.
- A "passed" verdict with no list of what was excluded.
- No record of provider or partner notices, although the quote promised them.
- Recommendations that would require a product you did not ask about.
- No retest route for any finding.
Ask for the report before you sign
Treat the deliverable as a quote line, not an afterthought. Before the test, ask the provider for a redacted sample report from a past engagement, the headings they will use, the delivery date, whether a readout meeting is included, who keeps the raw data and for how long, and whether a retest of fixed findings is priced in. The same logic applies to other services: the DDoS protection services quote checklist does it for the mitigation contract itself. If you cannot get a sample, treat that as information.
Limitations
This article is not legal or regulatory advice and does not state what any supervisor, auditor or licensing authority requires of a test report. It uses no external statistics or incident case studies. The matrix is a review checklist to adapt to your own platform, provider and contract, and it contains no thresholds. A stress test report describes how a tested path behaved under agreed traffic at one point in time; it does not cover every attack type, does not prove future uptime and is not a certification. A provider's behaviour can differ from its documentation, and a test environment may not mirror production exactly.
Decision and next step
If you can read the draft and, for each vector, find what reached the origin, what mitigation absorbed, how long it took and what real users felt, the report is doing its job. If you cannot, ask for the evidence behind the verdict or reject the deliverable. To have the report specified in the quote before any traffic is sent, scope an authorized test with AnySec and we will map your architecture to a test plan, the approvals to collect and the sections you will receive.
Sources and review
The framing follows public guidance: NIST SP 800-115 Technical Guide to Information Security Testing and Assessment, which covers planning, analysis and reporting of technical tests, and the Cloudflare Learning Center on DDoS attacks, read 2026-11-16; check the current versions, and read your own cloud and DDoS providers' current testing policies. AnySec prepared the report matrix from the public structure of authorised testing engagements; no AnySec-original statistics are used. Author: AnySec Engineering. Published 2026-11-16; reviewed by the AnySec security team.
Related reading
- DDoS stress testing: authorised test vs booter — what separates a sanctioned test from an attack.
- Scoping DDoS stress testing for an online casino — the scope the report has to reflect.
- What a real pentest report looks like — the same deliverable question for penetration testing.
To have the deliverable specified up front, get a fixed quote.
Frequently asked questions
- What should a DDoS stress test report contain?
- At minimum: the signed scope and rules of engagement, a timeline of what was sent and when, findings per attack vector showing what reached your origin and what the mitigation layer absorbed, how long detection and mitigation took to engage, the effect on legitimate traffic, a ranked list of fixes, and a statement of what was not tested. A peak-traffic screenshot and a pass or fail verdict are not enough on their own.
- Is a pass or fail verdict useful in a stress test report?
- Only as a summary of findings you can inspect. A verdict without the vector, the timeline and the observed behaviour cannot be challenged or repeated, so it gives a reviewer, auditor or board nothing to verify. Ask for the evidence behind every verdict and for the conditions under which the verdict would change.
- How long should a stress test report take to arrive?
- Agree the delivery date, and a short debrief, in the quote before the test. A report written weeks later from memory is weaker than one built from timestamps captured during the test. Many teams ask for a verbal readout within days and the written report after the logs from their own providers have been collected.
- Who should receive the report?
- The system owner who signed the authorisation, the person who owns the DDoS mitigation configuration, and whoever has to answer a reviewer or auditor. Treat it as sensitive: it describes how your defences behave under load, so restrict distribution and agree how long the provider keeps raw data.
- Do we need a retest after fixing the findings?
- If a finding led to a configuration change, a short retest of the affected vector is the only way to show the change worked. Ask whether a retest of fixed findings is included in the quote or priced separately, and what scope it covers.
Rather not learn this in production.
Talk to the engineers behind these write-ups — no sales script, a straight read on where you stand.
Get a fixed quote