The Cortex Query Language supports built-in datasets, custom datasets, and presets.
Every Cortex Query Language (XQL) dataset query begins by identifying a data source that the query will run against. Each data source has a unique name, and a series of fields. Your query specifies the data source, and then provides stages that identify fields of interest and perform operations against those fields.
You can query against either datasets or Presets in a dataset query. XQL supports using different languages for dataset and field names. In addition, the dataset formats supported are dependent on the data retention offerings available in Cortex XSIAM according to whether you want to query hot storage (default) or cold storage. For more information, see XQL Language Structure.
The standard, built-in data source that is available in every Cortex XSIAM instance is the xdr_data
dataset. This is a very large dataset with many available fields. For more information about this dataset, see Cortex XDR XQL Schema Reference. Cortex Query Language (XQL) supports using different languages for dataset and field names. In addition, the dataset formats supported are dependent on the data retention offerings available in Cortex XSIAM according to whether you want to query hot storage (default) or cold storage. For more information, see XQL Language Structure.
This dataset is comprised of both raw Endpoint Detection and Response (EDR) events reported by the Cortex XSIAM agent, and of logs from different sources such as third-party logs. To help you investigate events more efficiently, Cortex XSIAM also stitches these logs and events together into common schemas called stories. These stories are available using the Cortex XSIAM Presets.
When building queries in XQL, keep the following in mind with respect to datasets:
Use the
dataset
keyword to specify a dataset on your query.Create custom datasets using the target stage.
Dataset names can use uppercase characters, but in queries dataset names are always treated as if they are lowercase. In addition, dataset names are supported using different languages, numbers (
0-9
), and underscores (_
). Yet, underscores cannot be the first character of the name.Upon ingestion, all fields are retained even fields with a null value. You can also use XQL to query parsing rules for null values.
Depending on your integrations, you can have the following datasets available for queries:
Data | Dataset |
---|---|
Active Directory via Cloud Identity Engine |
NoteTo set up this Cloud Identity Engine (previously called Directory Sync Service (DSS)) dataset, you need to set up a Cloud Identity Engine. Otherwise, you will not have a |
Alerts table in Cortex XSIAM | alerts Note
|
Amazon S3 |
|
Authentication logs (subset of xdr_data) | Authentication logs, such as Okta: NoteThe fields contained in this dataset are a subset of the fields in the |
AWS CloudTrail and Amazon CloudWatch |
|
Azure Event Hub |
|
Azure Network Watcher |
|
BeyondTrust Privilege Management Cloud |
|
Box | Events (admin_logs)
Box Shield Alerts
Users
Groups
|
Checkpoint FW1/VPN1 |
|
Cisco ASA | Cisco ASA firewalls or Cisco AnyConnect VPN
|
Collector status change audit for collection integrations, custom collectors, and marketplace collectors. |
|
Corelight Zeek |
|
Correlation rule executions |
|
Cortex Data Lakes |
|
Cortex XDR Collectors |
|
Cortex XSIAM Host Firewall enforcement events |
|
CrowdStrike FDR |
|
CSV files in shared Windows directory | Custom datasets: Select from pre-existing user-created datasets or add a new dataset. |
Database data (MySQL, PostgreSQL, MSSQL, and Oracle) |
|
Data ingestion health metrics | Datasets:
Presets:
|
Dropbox | Events
Member Devices
Users
Groups
|
Elasticsearch Filebeat |
|
Elasticsearch Winlogbeat |
NoteIf the vendor and product are not specified in the Winlogbeat profile’s configuration file, Cortex XSIAM creates a default dataset called |
Errors related to Parsing Rules and Data Model Rules |
|
Errors related to event forwarding |
|
Forcepoint DLP |
|
Fortinet Fortigate |
|
GlobalProtect access authentication logs |
NoteTo ensure GlobalProtect access authentication logs are sent to Cortex XSIAM, verify that your PANW firewall’s Log Settings for GlobalProtect has the Cortex Data Lake checkbox selected. |
Google Cloud Platform (GCP) logs |
|
Google Kubernetes Engine (GKE) |
|
Google Workspace |
|
Host Inventory and Vulnerability Assessment |
|
Incidents table in Cortex XSIAM |
|
IT performance metrics |
|
JSON or text logs from third-party source over HTTP |
|
Login logs (subset of xdr_data) | Login logs, such as WEC: NoteThe fields contained in this dataset are a subset of the fields in the |
Logs from third party source over FTP, FTPS, or SFTP |
|
Microsoft Defender for Endpoint |
|
Microsoft 365 (email) |
|
Microsoft Office 365 |
|
NetFlow |
|
Network Share logs |
|
Okta |
|
OneLogin | Log collection
Directory
|
PANW EDR |
|
PANW IOT Security | Alerts
Devices
|
PANW NGFW |
Supports the following logs.
*These datasets use the query field names as described in the Cortex schema documentation. |
PingFederate |
|
PingOne for Enterprise |
|
Prisma Access Browser |
|
Prisma Cloud |
|
Prisma Cloud Compute |
|
Proofpoint Targeted Attack Protection |
|
SentinelOne DeepVisibility |
|
ServiceNow CMDB | A ServiceNow CMDB dataset is created for each table configured for data collection using the format |
Salesforce.com |
|
Syslog/CEF |
|
USB devices connect and disconnect events reported by the agent |
Note
|
VPN logs (subset of | VPN logs, such as GlobalProtect: NoteThe fields contained in this dataset are a subset of the fields in the |
Windows Endpoints using Cortex XDR Forensics Add-on |
|
Windows event logs via Cortex XDR Windows agents |
|
Windows Event Collector (WEC) |
|
Windows DHCP using Elasticsearch Filebeat |
|
Windows DNS Debug using Elasticsearch Filebeat | Raw Data
Normalized Stories
|
Workday |
|
Zscaler Cloud Firewall | ZIA
ZPA
|
Presets offer groupings of xdr_data
fields that are useful for analyzing specific areas of network and endpoint activity. All of the fields available for a preset are also available on the larger xdr_data
dataset, but by using the preset your query can run more efficiently. Presets are sorted at random by the first one million results found.
Two of the available presets are stories. These contain information stitched together from Cortex XSIAM agent events and log files to form a common schema. They are authentication_story
and network_story
.
You use the preset
keyword to specify a dataset in your query.