Detections and alerts

Use the Detections feature to create and manage rules, and view the alerts these rules create. Rules periodically search indices (such as endgame-* and filebeat-*) for suspicious source events, and create alerts when a rule’s conditions are met. When an alert is created, its status is Open. To help track investigations, an alert’s status can be set as Open, In progress, or Closed (see Change alert statuses).

In addition to creating your own rules, enable Elastic prebuilt rules to immediately start detecting suspicious activity. For detailed information on all the prebuilt rules, see the Prebuilt rule reference section. Once the prebuilt rules are loaded and running, Tune detection rules and Rule exceptions and value lists explain how to modify the rules to reduce false positives and get a better set of actionable alerts. You can also use exceptions and value lists when creating or modifying your own rules.

There are two special prebuilt rules you need to know about:

If you want to receive notifications via external systems, such as Slack or email, when alerts are created, use the Kibana Alerting and Actions framework.

After rules have started running, you can monitor their executions to verify they are functioning correctly, as well as view, manage, and troubleshoot alerts (see Managing detection alerts and Monitor and troubleshoot rule executions).

You can create and manage rules and alerts via the UI or the Detections API.

Cold tier is a data tier that holds time series data that is accessed only occasionally. In Elastic Stack version >=7.11.0, Elastic Security supports cold tier data for the following Elasticsearch indices:

Elastic Security does NOT support cold tier data for the following Elasticsearch indices:

Using cold tier data for unsupported indices may result in detection rule timeouts and overall performance degradation.

Indicator match rules provide a powerful capability to search your security data; however, their queries can consume significant deployment resources. As such, the following support restrictions are in place:

Detections prerequisites and requirements provides detailed information on all the permissions required to initiate and use the Detections feature.

Malware, short for malicious software, is any software program designed to damage or execute unauthorized actions on a computer system. Examples of malware include viruses, worms, Trojan horses, adware, scareware, and spyware. Some malware, such as viruses, can severely damage a computer’s hard drive by deleting files or directory information. Other malware, such as spyware, can obtain user data without their knowledge.

Malware may be stealthy and appear as legitimate executable code, scripts, active content, and other software. It is also often embedded in non-malicious files, non-suspicious websites, and standard programs — sometimes making the root source difficult to identify. If infected and not resolved promptly, malware can cause irreparable damage to a computer network.

For information on how to enable malware protection on your host, see Malware Protection.

To determine if a file is malicious or benign, a machine learning model looks for static attributes of files (without executing the file) that include file structure, layout, and content. This includes information such as file header data, imports, exports, section names, and file size. These attributes are extracted from millions of benign and malicious file samples, which then are passed to a machine-learning algorithm that distinguishes a benign file from a malicious one. The machine learning model is updated as new data is procured and analyzed.

A malware threshold determines the action the agent should take if malware is detected. The Elastic Agent uses a recommended threshold level that generates a balanced number of alerts with a low probability of undetected malware. This threshold also minimizes the number of false positive alerts.

Ransomware is computer malware that installs discreetly on a user’s computer and encrypts data until a specified amount of money (ransom) is paid. Ransomware is usually similar to other malware in its delivery and execution, infecting systems through spear-phishing or drive-by downloads. If not resolved immediately, ransomware can cause irreparable damage to an entire computer network.

Behavioral ransomware prevention on the Elastic Endpoint detects and stops ransomware attacks on Windows systems by analyzing data from low-level system processes, and is effective across an array of widespread ransomware families — including those targeting the system’s master boot record.

For information on how to enable ransomware protection on your host, see Ransomware protection.

Depending on your privileges and whether detection system indices have already been created for the Kibana space, you might see an error message when you try to open the Detections page.

Let’s set up your detection engine

If you see this message, a user with specific privileges must visit the Detections page before you can view detection rules and alerts. See Enable Detections for a list of all the requirements.

Detection engine permissions required

If you see this message, you do not have the required privileges to view the Detections page, and you should contact your Kibana administrator.