CrowdStrike process events (WIP)
Collects process execution events per endpoint from the CrowdStrike Falcon LogScale (Humio) API and stores them with the scan, one record per endpoint keyed by its Falcon device ID. It is a behavioural record of which executables ran on a host, rather than an inventory source.
| Type name | crowdstrike_process_events |
| Creates | Nothing — enriches endpoints that already exist |
| Runs on | A sensor with outbound reach to the Falcon API |
| Field prefix | None; events are stored as event data, not asset fields |
It defines no default fields and skips the identification stage entirely — it is an event store. Assets are not created, updated or deleted by it, so its action settings have no effect. Use CrowdStrike for inventory.
What it contributes
The sensor reads the endpoint list from this deployment's inventory, not from
Falcon. For every endpoint that has a cs_device_id (a 32-character Falcon device
ID; anything else is skipped), the integration runs a query job against the
Falcon investigate_view repository over the Start Window, and collects the
results grouped per host and executable image name.
That gives you, per asset, the distinct binaries that executed in the window. The query groups without counting, so it does not say how often each one ran. It is useful for spotting execution of things that have no business being on a host, and for understanding what a machine actually does before you decide how to treat it.
Requirements
The integration is outbound-only: it is a scheduled query client, so nothing is installed on the CrowdStrike side, nothing runs on the endpoints themselves, and nothing has to reach into the deployment.
Network access
| From | To | Protocol | Port |
|---|---|---|---|
| The sensor running the integration | The Falcon API host for your region, e.g. api.us-2.crowdstrike.com | HTTPS over TCP | 443 |
The token request and every query job go to that same host — the /humio path in
the API URL is a route on it, not a separate service — so one egress rule covers
the whole integration. Open it for your region's host, and if the sensor
reaches the internet through a proxy, point Proxy settings at it rather than
opening direct egress.
Access in CrowdStrike
| Need | Why |
|---|---|
| An account that can create API clients in the Falcon console | The credential is created under Support and resources → API clients and keys, which is admin-gated |
Query access to the LogScale investigate_view repository | The only Falcon API the run calls besides the token endpoint; the host list comes from the inventory, so no Hosts scope is used. See Permissions in CrowdStrike |
| A LogScale entitlement that permits query jobs | The events come from investigate_view via query jobs; whether that is available depends on the Falcon subscription |
In this deployment
| Need | Why |
|---|---|
| A registered, healthy sensor | The integration runs on a sensor, on the platform's schedule — see Sensors |
| Endpoints already discovered by CrowdStrike | One query job runs per known cs_device_id; with no CrowdStrike hosts in the inventory there is nothing to query for |
Because it issues one query job per endpoint, this is the longest-running integration in the catalogue on a large estate. Give it a sensor that stays up rather than the smallest one available, and read the cost warning under Configuration before widening the window.
Configuration
| Field | Required | Notes |
|---|---|---|
| Name | Yes | Label for this instance |
| API URL | Yes | The Falcon API base URL for your region, e.g. https://api.us-2.crowdstrike.com, without a path. The sensor appends /oauth2/token and /humio/api/v1/... itself |
| Client ID | Yes | Falcon API client ID |
| Client Secret | Yes | Falcon API client secret. Write-only: once saved it shows as a locked box with Replace, and is sent only to the sensor |
| Proxy settings | No | http://host:port |
| Sensor | Yes | Sensor that will run the integration |
| Ignore SSL | Yes | A checkbox. The collector does not currently apply it: certificates are always validated |
/humio off the API URLThe console's hint for this field shows an example ending in /humio. The
collector adds that path itself, so a URL ending in /humio sends the token
request to the wrong address and authentication fails.
Test connection requests a token and then submits a one-minute,
single-event query job to investigate_view, cancelling it straight away. It
therefore proves query access, not just the credentials.
Unlike other integrations, its advanced settings are specific rather than generic:
| Setting | Default | What it does |
|---|---|---|
| Start Window | 7d | How far back to look for events |
| Query Timeout (sec) | 120 | Maximum wait for a single query job |
| Poll Interval (sec) | 2 | Delay between polls of a query job |
| Max Lines / Host | 1000 | Cap on events collected per host |
One query job runs per endpoint. A wide start window across a large estate is a lot of LogScale work and a long run. Start at the defaults and widen only if the seven-day view is genuinely insufficient.
Permissions in CrowdStrike
The API client needs read access to the event search / LogScale repository the
queries target (investigate_view). No Hosts scope is needed: the endpoints come
from the inventory. When access is missing, Test connection names the
permission as Investigate (LogScale): Read.
Confirm the exact Falcon API scope name that grants query-job access to the
investigate_view repository for the current Falcon console, and document
whether an on-demand LogScale subscription is a prerequisite.
See also
- CrowdStrike — the inventory source this depends on
- Integrations overview — the shared configuration model