Want to learn more?
We’re here to answer any questions you have!
Take a tour
TRUSTED BY COMPANIES WORLDWIDE
Test system resilience proactively with controlled experiments
Steadybit is reliability platform that helps teams assess and improve the resilience of their services. With automated issue discovery and controlled experiments, you can find and validate system weaknesses before they become outages.
Unlike other chaos engineering tools, Steadybit uses an open source extension framework to quickly connect to popular tools across your tech stack. Need a custom integration? You can easily add it yourself using our Extension Kits.
We’ve supported both SaaS and On-Prem deployments since Day 1.
Validate Monitoring Alerts
Run scenarios to check your alert coverage and accuracy
Reduce Reliability Risks
Catch reliability issues and fix them before they reach production
Resolve Incidents Faster
Train your team to be able to handle any incident quickly
Build experiments with no-code actions & templates
Drag-and-drop actions into the Steadybit experiment editor to create new reliability tests and iterate quickly.
Network
Kubernetes
Cloud Services
Physical & Virtual Hosts
Applications
Observability
Blackhole Subnet Attack
This action lets you simulate an outage by assigning a network ACL to a selected subnet.
Blackhole Zone Attack
Simulate an outage of an entire availability zone to see whether traffic is rerouted correctly.
Corrupt Outgoing Packages
Inject corrupt packets by introducing single bit error at a random offset into egress traffic.
Drop Outgoing Traffic
This action causes packet loss for outgoing network traff (egress).
Block DNS
This network attack blocks access to DNS servers for a specific duration.
Block Traffic
This action drops all traffic to specified hostnames, IP addresses, and ports.
Delay Outgoing Traffic
Use this action to inject latency into egress network traffic for a certain duration.Explore the Action Library
Browse open source actions that you can easily add to experiments.
Delete Pod
This attack allows you to delete one or multiple pods to test the resilience of your application.
Cause Crash Loop
This action continuously kills specified containers in a selected pod.
Rollout Restart Deployment
Simulate the rollout of a Kubernetes deployment using a kubectl command.
Pause Docker Container
Run this action to pause one or more containers for a certain amount of time.
Taint a Node
Use this attack to taint one or multiple nodes for a given duration.
Drain Node
Use this attack to drain one or multiple nodes and check performance degradation.
Stop Container
Check the exit behavior and restart process by terminating one or more containers.Explore the Action Library
Browse open source actions that you can easily add to experiments.
Change Azure VM State
This action allows you to reboot, delete, stop or deallocate Azure virtual machines.
Change EC2 Instance State
Reboot, stop, hibernate and terminate EC2 instances during an experiment.
Change GCP VM State
Reset, delete, stop or suspend GCP virtual machines during an experiment.
Run AWS FIS Experiment
Execute AWS FIS Experiments via Steadybit to manage everything in one place.
Trigger DB Instance Stop
Test disaster recovery processes by stopping RDS database instances.
Reboot RDS Instances
This action enables you to reboot a single RDS database instance.
Trigger DB Cluster Failover
This action triggers DB cluster failover by promoting a standby instance to primary.Explore the Action Library
Browse open source actions that you can easily add to experiments.
Stress CPU
Test your application's resilience to high CPU load by generating load for one or more cores.
Stress IO
Generate read/write operations on hard disks or ephemeral storage for a given duration.
Stress Memory
Stress a specific amount of memory using ongoing memory allocations, reads and writes.
Trigger Shutdown Host
This action triggers a reboot or shutdown of the host to validate failover processes and impact.
Fill Disk
This action fills the container's ephemeral storage with random data for a given duration.
Time Travel
Test your application's ability to handle time changes by changing the clock time.
Change CPU Frequency
Dynamically adjust the CPU frequency limits across all cores for a specified duration.Explore the Action Library
Browse open source actions that you can easily add to experiments.
Inject Latency
Use this action to inject latency into AWS Lambda or Azure functions.
Inject Exception
This action injects exceptions into applications for a set amount of time.
Inject Status Code
Inject a fixed status code to test how upstream services respond to specific HTTP statuses.
Inject Controller Exception
Inject a RuntimeException into a Spring™ MVC controller before the handler method is executed.
Inject Java Method Exception
Inject a RuntimeException into a public Java method for a given amount of time.
Java Method Delay
Run this attack to inject latency into any Java-based application for a given duration.
Fill Diskspace
This action fills the temporary disk space of on AWS Lambda or Azure function.Explore the Action Library
Browse open source actions that you can easily add to experiments.
Create Maintenance Window
Create a maintenance window to avoid false positives in your monitoring system.
Check Monitor Status
This action collects information about a specified monitor and verifies an expected status.
Create Monitor Downtime
Mute Datadog monitors during experiments to not create unnecessary noise.
Check Grafana Alert Rule State
Collect information about the state of the Grafana alert rules during an experiment.
Gather Prometheus Metrics
Collect Prometheus metrics during an experiment to help validate your hypothesis.
Check SLO State in Splunk
Collect information on the SLO state in Splunk so you can check application performance.
Create Muting Rule in New Relic
Mute alerts for a specified amount of time so experiments don't create extra noise.Explore the Action Library
Browse open source actions that you can easily add to experiments.
Explore and select targets for experiments
When you install our agent on your network, Steadybit will automatically discover any potential experiment targets and pull in related metadata from your testing environment. Our intuitive query language makes it easy to group and filter your targets however you want.
Get advice on what experiments to run first
To help you get started fast, our Reliability Advice feature will provide you with insights on if there are any common reliability issues detected.
You’ll see instructions on how to fix any issues in your code, and then we’ll recommend which experiments would be valuable to run next.
Design, customize, and run experiments
Design full experiments in seconds using templates for popular use cases and our drag-and-drop editor. With our open source framework, you can easily add custom actions and extensions to run any type of experiment you want.
Once you’re happy with an experiment, you can automate your test executions with the Steadybit API or CLI.
Why SRE & platform teams choose us
Our customers inspire us everyday with new experiment types and custom extensions to really push their systems to the limit.
-
Krishna Palati
Director of Software Engineering
-
Jan Rundshagen
Cloud Platform Engineer
-
Ilias Tsakiridis
Site Reliability Engineering Team Lead
Shift to a proactive reliability approach
Measure your current reliability posture and validate system behaviors with controlled tests.
Reliability Advice
Automatically detect vulnerabilities
Assess whether your targets are compliant with reliability best practices.
Experiment Editor
Run actions with a timeline-based editor
Start quick with templates for common use cases or build fully custom tests.
Assign Teams & Roles
Set guardrails & fine-grained permissions
Define access and permissions for users to ensure safe testing.
Extend Steadybit to perfectly fit your systems
To get started, you will need to install the Steadybit agent on your network and add any of our open source extensions that match your tech stack. Then, you can use the Steadybit platform to view targets, design experiments, and run tests.
FAQs
Evaluating chaos engineering tools? Here are the most common questions we get from teams.
Can we deploy Steadybit in On-Prem or air-gapped environments?
Yes, of course! From Day 1, Steadybit has offered SaaS and On-Prem deployment options with full feature parity. Install the control plane and extensions in any environment seamlessly and start improving your reliability.
To learn more about our On-Prem support, you can read the installation details here.
How can we evaluate Steadybit to see if it's right for us?
If you’re not sure the best way to get started, a quick call with us can be helpful. We can answer technical questions you have and guide you on what we’ve seen work the best. You can schedule that here.
If you want to get into the platform and start playing around right away, we offer a free 30-day trial. You can either install agents and extensions directly on your systems or use our provided sample data to see how each of our features work. Sign up here.
If none of these sound right, just fill out our contact us form and provide us with more info. We’re here to help!
How do we add custom actions and extensions?
Steadybit is the most extensible reliability platform because it has a hybrid architecture that supports open source extensions.
Our ExtensionKits enable you to add custom actions, templates, targets, advice, and extensions. Write in your preferred coding language and start to customize Steadybit to fit your specific use cases and tech stack.
How does Steadybit automatically detect reliability vulnerabilities?
Our Reliability Advice feature continually analyzes all of your discovered targets and checks whether they are compliant with the best practices outlined in the “Advice” settings.
When you get started with Steadybit, there are 13 Advice checks out-of-the-box based on the best practices outlined by the open source tool, kube-score.
If you want to add checks based on internal standards or other best practices, our AdviceKit provides instructions on how to write your own custom Advice.
What prevents experiments from causing unintended damage?
To start, we have RBAC user permissions that let you limit the actions and targets that users can interact with. Group targets into defined testing environments and assign only the relevant teams to ensure least privilege access.
When designing experiments, you can select a blast radius for your targets. For example, you could specify that you only want to target 10% of the pods in a cluster. This is an easy way to ensure that your experiments start small with limited impact.
Before an experiment runs, you can configure pre-flight webhooks. These customizable checks allow you to ensure that all conditions are ready for your experiment to begin running.
When experiments are running, anyone in your organization is able to hit the “Emergency Stop” button. This will immediately rollback changes and ensure that you can respond fast.
With all of the features, you can set up controls and guardrails to enable experimenting with confidence.
Get a Personalized Demo
Ready to hear more about Steadybit?
Schedule a demo with our team to see a platform walk-through and get your questions answered.
