Generate a variety of suspect actions that are detected by Falco rulesets.
Warning — We strongly recommend that you run the program within Docker (see below), since some commands might alter your system. For example, some actions modify files and directories below /bin, /etc, /dev, etc. Make sure you fully understand what is the purpose of this tool before running any action.
Notice — From version v0.11.0 the event-generator requires Falco 0.37.0 or newer. Previous versions of the event-generator might be compatible with older versions of Falco, however, we do not guarantee it.
The full command line documentation is here.
$ event-generator list --all
helper.ExecLs
helper.NetworkActivity
helper.RunShell
k8saudit.ClusterRoleWithPodExecCreated
k8saudit.ClusterRoleWithWildcardCreated
k8saudit.ClusterRoleWithWritePrivilegesCreated
k8saudit.CreateDisallowedPod
k8saudit.CreateHostNetworkPod
k8saudit.CreateModifyConfigmapWithPrivateCredentials
k8saudit.CreateNodePortService
k8saudit.CreatePrivilegedPod
k8saudit.CreateSensitiveMountPod
k8saudit.K8SConfigMapCreated
k8saudit.K8SDeploymentCreated
k8saudit.K8SServiceCreated
k8saudit.K8SServiceaccountCreated
syscall.ChangeThreadNamespace
syscall.CreateFilesBelowDev
syscall.CreateSymlinkOverSensitiveFiles
syscall.DbProgramSpawnedProcess
syscall.DirectoryTraversalMonitoredFileRead
syscall.MkdirBinaryDirs
syscall.ModifyBinaryDirs
syscall.NonSudoSetuid
syscall.ReadSensitiveFileTrustedAfterStartup
syscall.ReadSensitiveFileUntrusted
syscall.RunShellUntrusted
syscall.ScheduleCronJobs
syscall.SearchPrivateKeysOrPasswords
syscall.SystemProcsNetworkActivity
syscall.SystemUserInteractive
syscall.UserMgmtBinaries
syscall.WriteBelowBinaryDir
syscall.WriteBelowEtc
syscall.WriteBelowRpmDatabaseevent-generator run [regexp]
Without arguments, it runs all actions; otherwise, only those actions match the given regular expression.
For example, to run only those actions containing the word Files in their name:
$ sudo event-generator run syscall\.\*Files\.\*
INFO sleep for 100ms action=syscall.ReadSensitiveFileUntrusted
INFO action executed action=syscall.ReadSensitiveFileUntrusted
INFO sleep for 100ms action=syscall.CreateSymlinkOverSensitiveFiles
INFO action executed action=syscall.CreateSymlinkOverSensitiveFiles
INFO sleep for 100ms action=syscall.DirectoryTraversalMonitoredFileRead
INFO action executed action=syscall.DirectoryTraversalMonitoredFileRead
INFO sleep for 100ms action=syscall.ReadSensitiveFileTrustedAfterStartup
INFO spawn as "httpd" action=syscall.ReadSensitiveFileTrustedAfterStartup args="^syscall.ReadSensitiveFileUntrusted$ --sleep 6s"
INFO sleep for 6s action=syscall.ReadSensitiveFileUntrusted as=httpd
INFO action executed action=syscall.ReadSensitiveFileUntrusted as=httpdUseful options:
--loopto run actions in a loop--sleepto set the length of time to wait before running an action (default to1s)
Also, note that not all actions are enabled by default. To run all actions, use the --all option.
Further options are documented here.
Run all events with the Docker image locally:
docker run -it --rm falcosecurity/event-generator runIt can be deployed in a Kubernetes cluster using the event-generator helm chart.
Before installing the chart, add the falcosecurity charts repository:
helm repo add falcosecurity https://falcosecurity.github.io/charts
helm repo updateRun all events once using a Kubernetes job:
helm install event-generator falcosecurity/event-generator \
--namespace event-generator \
--create-namespace \
--set config.loop=false \
--set config.actions=""Run all events in a loop using a Kubernetes deployment:
helm install event-generator falcosecurity/event-generator \
--namespace event-generator \
--create-namespace \
--set config.actions=""N.B.
The above commands apply to the event-generator namespace. Use a different name to use a different namespace. It will generate events in the same namespace.
The syscall collection performs a variety of suspect actions detected by the default Falco ruleset.
$ docker run -it --rm falcosecurity/event-generator run syscall --loopThe above command loops forever, incessantly generating a sample event each second.
The k8saudit collection generates activity that matches the k8s audit event ruleset.
Note that all k8saudit are disabled by default. To enable them, use the --all option.
$ event-generator run k8saudit --all --loop --namespace `falco-eg-sandbox`N.B.: the namespace must exist already.
The above command loops forever, creating resources in the falco-eg-sandbox namespace and deleting the after each iteration.
N.B.
- the namespace must already exist
- to produce any effect the Kubernetes audit log must be enabled, see here
Since v0.4.0, this tool introduces a convenient integration test suite for Falco rules. The event-generator test command can run actions and test them against a running Falco instance.
This feature requires Falco 0.24.0 or newer. Before using the command below, you need Falco installed and running with the gRPC Output enabled.
Run the following command to test syscall actions on a local Falco instance (connects via Unix socket to /run/falco/falco.sock by default):
sudo ./event-generator test syscallBefore running the following commands make sure you have added the falcosecurity charts repository as explained here.
Test all events once using a Kubernetes job:
helm install event-generator falcosecurity/event-generator \
--namespace event-generator \
--create-namespace \
--set config.command=test \
--set config.loop=false \
--set config.actions=""Test all events in a loop using a Kubernetes deployment:
helm install event-generator falcosecurity/event-generator \
--namespace event-generator \
--create-namespace \
--set config.command=test \
--set config.actions=""Note that to test k8saudit events, you need Kubernetes Audit Log functionality enabled in Kubernetes and the k8saudit plugin in Falco.
Since v0.5.0, the event-generator can also be used for benchmarking a running instance of Falco. The command event-generator bench generates a high number of Event Per Second (EPS) to show you events throughput allowed by your Falco installation.
Be aware that before Falco 0.37 a rate-limiter for notifications that affects the gRPC Outputs APIs was present. You probably need to increase the outputs.rate and outputs.max_burst values within the Falco configuration, otherwise EPS will be rate-limited by the throttling mechanism.
Before starting a benchmark, the most important thing to understand is that the --sleep option controls the number of EPS (default to 250ms): reducing this value will increase the EPS. Furthermore, if the --loop option is set, the sleeping duration is automatically halved on each round. The --pid option can be used to monitor the Falco process.
You can find more details about the command-line usage here.
Please, keep in mind that not all actions can be used for benchmarking since some of them take too long to generate a high number of EPS. For example, k8saudit actions are not supposed to work, since those actions need some time to create Kubernetes resources. Also, some syscall actions sleep for a while (like the syscall.ReadSensitiveFileUntrusted) thus cannot be used.
Benchmark example
A common way for benchmarking a local Falco instance is by running the following command (that connects via Unix socket to /run/falco/falco.sock by default):
sudo event-generator bench "ChangeThreadNamespace|ReadSensitiveFileUntrusted" --all --loop --sleep 10ms --pid $(pidof -s falco)See the events registry.
Sure!
Check out the events registry conventions, then feel free to open a PR!
Your contribution is highly appreciated.
This project provides three main packages that can be imported and used separately:
/cmdcontains the CLI implementation/eventscontains the events registry/pkg/runnercontains the actions runner implementations
Feel free to use them as you like on your projects.
Special thanks to Mark Stemm (@mstemm) — the author of the first event generator.