You want to know how many users your web application can handle or how it behaves under a DDoS attack? We simulate the agreed load over the internet and accompany the tests personally.
HTTP/HTTPS scenarios over the internet, personally accompanied test runs and repeatable load parameters.
DDoS simulation: knowing how much of a storm your systems can weather.
Your situation
You want to know whether your web application can withstand the expected rush – or a DDoS attack?
As the operator of a web application, you face one of three questions: can the application be provided with good performance for a defined number of users whose behaviour you can estimate? At what number of users does performance drop? And can you defend yourself if you are attacked by many computers at the same time (distributed denial of service)?
Load tests are often run internally for this. However, this does not test the actual target structure, because real traffic comes over the internet. Maintaining your own infrastructure for realistic tests is costly.
We use the infrastructure of cloud providers and test your scenarios under real conditions: at application level, with real HTTP or HTTPS connections and, on request, from the USA, Asia or South America. Our testers support the tests personally – as part of our penetration testing portfolio.
Packages
Test packages for DDoS simulation and load tests
One test scenario corresponds to four hours of effective testing time. The package prices (net) include set-up time, documentation, test execution and infrastructure costs. “Up to” means that the stated number of source addresses or sessions can be used within the booked package at no extra charge.
Package sizes and prices (net)
Small: up to 5 source addresses, up to 250,000 maximum sessions, up to 10 URLs – €5,500
Medium: up to 10 source addresses, up to 500,000 maximum sessions, up to 50 URLs – €6,000
Large: up to 50 source addresses, up to 2,500,000 maximum sessions, unlimited URLs, test from the US, Asia or South America – €8,500
More: any number of source addresses and sessions, unlimited URLs, test from the US, Asia or South America – project-specific
Included in every package
Personal support and advice before, during and after the tests
Individual report
One machine per IP address with 1 Gbit/s bandwidth
Different user agents within one test on request
URLs with HTTP GET and HTTP POST (POST parameters definable)
Set-up time, documentation, test execution and infrastructure costs
Services & results
Sizing, stress, DDoS: define the scenario, increase the load, with personal support
You see how your application reacts under load: automated test users call agreed HTTP/HTTPS scenarios over the internet, as a sizing test, stress test or DDoS simulation with increasing load. One test scenario comprises four hours of effective testing time; our testers accompany every run personally and can stop it at any time.
To obtain a meaningful result, we test at application level. Unlike tests that aim to saturate bandwidth, we define real HTTP/HTTPS connections as scenarios, which are called by a set number of automated test users. Usually only a fraction of the server-side bandwidth is needed – it is the front-end servers and back-end systems that are put under heavy load.
01
Choosing the test objective
First, we clarify which question your operations need to answer:
Sizing test: can the application be provided with good performance for a defined number of users?
Stress test: at what number of users does performance drop?
DDoS simulation: does the application withstand a distributed attack from many computers?
02
Defining the test scenario
For each test scenario, we define the parameters together. One test scenario corresponds to four hours of effective testing time, during which we vary the test duration and the number of target sessions.
URLs to be called, both HTTP GET and HTTP POST (POST parameters definable)
Number of source IP addresses
Duration of the test runs
Number of target sessions
Different user agents within one test on request
03
Agreeing the test period, increasing the load step by step
As a rule, we carry out the tests several times, with the load increasing with each test run. This allows us to identify performance problems quickly and narrow them down. We use one machine with 1 Gbit/s bandwidth per IP address.
Because identical scenarios can be repeated, you can also use them to test the effect of configuration changes, for example to load balancer or server parameters.
04
Personally supported and can be stopped at any time
The risk of these tests is an outage – in stress tests and DDoS simulations, it is even the objective. Our testers therefore support the tests personally and can stop them at any time. In a real overload situation, you cannot.
In most cases, the systems recover on their own; sometimes a restart is necessary. We agree on stopping and restarting with the people responsible for operations in advance.
05
Report and advice
We support and advise you before, during and after the tests. Set-up time, documentation, test execution and infrastructure costs are included in the package price.
Your individual report
An individually prepared report on the test scenarios carried out – as a basis for narrowing down performance problems and optimising configurations.
Our expertise
DDoS simulation, sizing and stress tests are part of our penetration testing portfolio. secuvera has been carrying out penetration tests since 2000 and is a BSI-certified IT security service provider for penetration testing.
The difference from an internal load test: we test via the infrastructure of cloud providers and therefore via the same route as your real users – the internet. On request, the traffic comes from the USA, Asia or South America.
Every test is personally accompanied by our testers. That is the decisive difference from a real overload situation: we can stop at any time.
Getting started
Defining user behaviour, load target and test window
Choose the test objective
Is it about a defined number of users, the performance limit or resilience against distributed overload? We determine whether a sizing test, stress test or DDoS simulation fits your question.
Agree on the test window and operations contacts
We agree on the test period, load parameters and contacts. Since outages are possible, stopping and restarting must also be clarified before the test.
Load target and test window chosen so that operations keep running.
Preparing for outages and test termination
Performance drops and outages are possible in stress and DDoS tests. Test period, authorisations and termination conditions are agreed before execution. Testing is carried out primarily at application level via HTTP/HTTPS.
Questions about DDoS simulations, sizing and stress tests
Which test fits: sizing test, stress test or DDoS simulation?
A sizing test focuses on a defined number of users. A stress test examines the performance limit; a DDoS simulation looks at distributed overload in the agreed scenario. We choose the approach based on the question your operations need to answer.
What do we need to define for a load scenario?
We need the relevant application steps and load parameters, such as the number of automated test users. In addition, the test period, permitted targets and contacts in operations. As a rule, the load is increased over several test runs.
Can a stress test or DDoS simulation lead to outages?
Yes. In stress tests and DDoS simulations, an outage is even the objective of the test. Our testers accompany the runs personally and can stop them at any time. In most cases, the systems recover on their own; sometimes a restart is necessary. We agree on stopping and restarting with the people responsible for operations before we start.
How long does a test scenario take?
One test scenario corresponds to four hours of effective testing time. During this time, we vary the test duration and the number of target sessions and carry out the test runs with increasing load.
How much does a DDoS simulation or load test cost?
We offer packages at net prices: Small €5,500, Medium €6,000, Large €8,500; we calculate larger scopes on a project-specific basis. The packages differ in the number of source addresses, maximum sessions and URLs. Set-up time, documentation, test execution and infrastructure costs are included.