---
sourceDocument: Zurich IT Operations Management
sourceDocumentLink: https://www.servicenow.com/docs/r/zurich/it-operations-management

 Release :

    - zurich

ft:locale :

    - en-US

ft:publication_title :

    - Zurich IT Operations Management

ft:clusterId :

    - itom

bundleId :

    - itom

workflow :

    - Technology


---

# Alert grouping and use cases

# Alert grouping and use cases {#ariaid-title1}

Release version: Zurich  
Updated July 31, 2025  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 6 minutes to read
Summarize  
![AI sparkle icon](https://servicenow.com/docs/portal-asset/ai-sparkle-icon) Summarized using AI  
This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.  

## Summary of Alert grouping and use cases

Alert grouping in ServiceNow enables IT teams to organize related alerts into consolidated groups, improving incident management and reducing alert noise.
Various grouping methods are available, ranging from manual and rule-based approaches to advanced, AI-driven algorithms.
Each method addresses specific use cases and environments, helping customers efficiently detect, correlate, and respond to issues across their IT infrastructure.
Show full answer Show less  

## Alert Grouping Methods and Use Cases

* **Log Analytics Grouping:** Uses log data analysis to correlate alerts based on log patterns and sequences, enabling detection of complex multi-step issues. Ideal for proactive monitoring, such as predicting server crashes by identifying error patterns in logs.
* **Rule-based Grouping:** Groups alerts according to user-defined rules and conditions, suitable for consistent and repeatable alert patterns. Commonly applied in scenarios like managing server issues during high-traffic events by linking parent server alerts with related symptoms.
* **Automated Grouping:** Employs machine learning and AI algorithms to detect patterns in historical alert data and automatically group related alerts. This dynamic method adapts to new issues, enhancing response times in complex environments like financial institutions managing numerous servers.
* **Mixed Grouping:** Combines multiple grouping strategies, such as CMDB-based and tag-based grouping, to leverage their strengths. Useful when grouping alerts within CMDB relationships or using tags to cluster alerts when a CMDB is unavailable, facilitating efficient root cause analysis and issue resolution.
* **Network-Traffic Based Grouping:** Analyzes network connections between processes, leveraging ML Service Mapping data to group alerts related to network traffic problems. This method helps pinpoint root causes in interconnected services, improving troubleshooting of issues like transaction slowdowns.
* **Text-based Grouping:** Utilizes natural language processing to group alerts by similarities in alert descriptions and attributes, effective for unstructured data. It supports grouping related alerts when CMDB data is unavailable, aiding faster identification of root causes in scenarios such as multiple similar device outages.
* **Manual Grouping:** Allows users to manually select and group related alerts based on their expertise, providing precise control but requiring more effort. Useful when automated methods are insufficient or when specialized knowledge is needed, such as grouping alerts linked to a single hardware failure.

## Parent Alert Selection

The parent alert in a group is either a real alert or a virtual alert depending on the grouping method. Manual, rule-based, and log analytics grouping designate a real alert as the parent, while methods like automatic, CMDB, text-based, tag cluster, and network traffic grouping create a virtual alert representing the oldest and most severe alert. By default, the oldest and highest severity alert becomes the main alert. Customization of this logic is possible through the **GroupingProduceVirtualHook** script include and the **getVirtual** method.

## Additional Considerations

* Alert groups are domain-specific in domain-separated environments, ensuring isolation of alerts within the same domain.
* Scheduled jobs run regularly (daily and every minute) to process alerts and generate groups based on defined patterns and relationships.
* ServiceNow documentation provides detailed configuration options for alert grouping, allowing customers to tailor grouping strategies to their operational needs.  
Alert grouping methods range from user-defined approaches, like Manual and Rule-based to advanced, fine-tunable algorithms, including
Automatic, Mixed, Text-based, Log Analytics, and Network Traffic-based grouping.  
{#alert-group-use-cases__table_dwb_r55_hcc__entry__3}

| Type | Description | Use case |
|-|-|-|
| Log Analytics Grouping | Alerts are grouped based on the analysis of log data. This involves correlating log entries to identify related alerts and issues. By leveraging log patterns and sequences, this method can detect complex, multi-step problems across the IT environment. | An online gaming company enhances server stability by implementing proactive log analytics. They monitor logs from game servers in real-time and use analysis tools to detect patterns of errors that occur before crashes. For instance, the analysis reveals that certain error patterns appear about 30 minutes prior to server crashes. By setting up automated alerts for these patterns, the company can initiate remediation actions, such as restarting services or reallocating resources, before a crash occurs. This proactive approach prevents disruptions, minimizes downtime, and improves the gaming experience by addressing issues before they impact players. |
| Rule-based Grouping | Alerts are grouped according to predefined rules and criteria set by users. These rules might include specific conditions, such as thresholds or event types. This method is effective for consistent and repeatable patterns but requires maintenance of the rules. | In a data center managing an e-commerce website, rule-based alert grouping helps handle high traffic during events like flash sales. Alerts about server issues, such as high CPU usage, are designated as parent alerts. These parent alerts are linked to child alerts that report related problems, like slow database queries. The rules ensure that server-related alerts are grouped with their symptoms, allowing the IT team to quickly identify and address server overload issues. This approach improves issue resolution efficiency and minimizes downtime. |
| Automated Grouping | Advanced algorithms automatically identify and group related alerts based on patterns and similarities in the alert data. This method leverages machine learning and AI to adapt to new and unknown issues, providing proactive alert management. Event Management groups alerts that are similar, but not necessarily identical, based on the proximity in time of the last event generation. Alerts with the same CI and the same pattern identifier are grouped together. Automatic alert grouping consists of the following components. * Alert Aggregation Learner (Service Analytics Alert Aggregation Learner - Daily): This offline job runs daily to process past alerts and perform statistical analysis to build alert patterns. For details, see [Configure pattern based alert grouping](https://www.servicenow.com/docs/~94ORiFmqC6cBTXZkKVscQ "Configure the Alert Aggregation Learner (Service Analytics Alert Aggregation Learner - Daily), an offline job that runs daily to process past alerts. It identifies patterns of related alerts using a combination of pattern-based and probabilistic techniques, enabling quicker detection and resolution of recurring issues."). * Real-time Alert Aggregation job (Service Analytics group alerts using RCA/Alert Aggregation): This job runs every minute to generate alert aggregation groups based on alert patterns, CMDB relationships, text similarity, user-defined alert clustering tags, and network traffic connection between processes. {#alert-group-use-cases__ul_jyj_qn2_jcc} | A large financial institution uses machine learning to manage alerts from numerous servers and applications. The system analyzes historical alert data to recognize patterns, such as database server failures frequently being accompanied by client connection errors. It then automatically groups related alerts together. For instance, when a new database server failure alert is detected, it is grouped with previous connection error alerts. This automated grouping helps the IT and security teams quickly identify and address issues, improving response times and reducing downtime. |
| Mixed Grouping | Mixed Grouping method combines alerts using multiple grouping strategies, such as CMDB-based grouping and tag-based grouping, into a single, cohesive group. It leverages the strengths of each strategy to reduce alert noise, improve alert correlation, and highlight the true root cause of incidents. * CMDB-based Grouping: Alerts are grouped based on Configuration Item (CI) relationships and dependencies from the Configuration Management Database (CMDB). This approach ensures that alerts related to specific infrastructure components or services are grouped together, providing context-aware alert management. * Tag Cluster Grouping: Alerts are categorized and grouped using tags or labels that represent common attributes, such as application, server type, or geographic location. This method allows for flexible and dynamic grouping based on evolving tagging strategies. {#alert-group-use-cases__ul_wks_dvc_kfc} | Use case for CMDB-based grouping: A network team uses CMDB data to group alerts for a specific router and its connected devices based on CMDB relationships. All related issues are visible in one place for root cause analysis. Use case for tag cluster grouping: An organization without a CMDB manages a Linux server running various services. The IT team uses the Node field in each alert to identify the server and group related events by node value. For example, alerts such as Service A down and Service B high CPU usage are clustered together when they share the same node value. This approach helps the IT team address server-related issues more efficiently. By clustering alerts for the same node, application, or IP address, the team streamlines their response efforts and resolves issues more effectively, even without a CMDB. |
| Network-traffic based Grouping | Network-traffic-based alert grouping analyzes network connections between processes across hosts to identify related alerts. This method leverages service candidates detected through ML Service Mapping, ensuring that alerts related to network traffic issues are grouped together for better context and faster alert resolution. | A cloud-based e-commerce platform experiences transaction slowdowns, causing delays in payment processing. Traditional alerting generates separate alerts for API timeouts, database lags, and network issues, making it difficult to pinpoint the root cause. With Network-Traffic Based Grouping, alerts are automatically grouped based on process-to-process connections identified through ML Service Mapping. The system detects that payment gateway services, fraud detection, and order processing are part of the same service candidate. This reveals that an overloaded fraud detection process is causing transaction bottlenecks. By scaling up the service, the team quickly resolves the issue, minimizing downtime and improving customer experience. |
| Text-based Grouping | Alerts are grouped by analyzing the text content of alerts to identify similarities and related issues. Natural language processing (NLP) techniques are used to find commonalities in alert description, metric name, and ci class, making this method effective for unstructured data. | When a Zoom room server experiences an outage, the IT team receives many alerts --- one per affected room. Each alert differs only by room number, for example, Zoom room 10 is down or Zoom room 11 is down. With a CMDB, alerts are grouped using CMDB relations based on the server's impact on associated Zoom rooms. Without a CMDB, text-based grouping uses natural language processing to group alerts with similar descriptions, helping the IT team identify and address the root cause faster. |
| Manual Grouping | Users manually select and group related alerts based on their expertise and understanding of the system. This approach allows for precise control but can be time-consuming and may miss automated correlations. | A system administrator receives multiple alerts about different services failing on a single server. The admin manually groups these alerts, identifying them as related to a single hardware failure on that server. The admin then prioritizes fixing the hardware issue to restore all services. |
[Table 1. Alert grouping types and use cases]

{#alert-group-use-cases__table_dwb_r55_hcc}

Grouping methods differ in how the parent alert is selected. In manual, rule-based, or log analytics grouping, a real alert is designated as the parent. In automatic, CMDB, text-based, tag cluster, and network traffic grouping, a
virtual alert representing the oldest and most severe alert in the group is created as the parent.  
By default, the main alert of the group is the oldest and the alert with the highest severity. There is a script include, GroupingProduceVirtualHook, using which we can change the main alert selection logic. The method that you should customize is getVirtual.  
Note:  
In domain-separated environments, alert groups are created only for alerts within the same domain.

For information on scheduled jobs and parameters, refer to [Scheduled jobs and parameters for alert grouping](https://www.servicenow.com/docs/da3Z3NXFhymwVM47uEoBkA "Automate alert organization by configuring jobs to group alerts based on predefined criteria and parameters."). For detailed information on different grouping types, see [Alert grouping types and creation methods](https://www.servicenow.com/docs/oi89wQZ9FdN0r3W6XxZCmw "Explore different alert grouping types, understand their descriptions, and learn about their creation methods to enhance problem identification and streamline alert management.").

