--- sourceDocument: Australia IT Operations Management sourceDocumentLink: https://www.servicenow.com/docs/r/australia/it-operations-management Release : - australia ft:locale : - en-US ft:publication_title : - Australia IT Operations Management ft:clusterId : - itom bundleId : - itom workflow : - Technology --- # Pattern Diagnostic agentic workflow # Pattern Diagnostic agentic workflow {#ariaid-title1} Release version: Australia Updated September 1, 2026 ![](https://www.servicenow.com/docs/portal-asset/ico-clock) 4 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 Pattern Diagnostic agentic workflow The Pattern Diagnostic agentic workflow streamlines the investigation of missing Configuration Item (CI) attributes discovered via pattern-based Discovery in ServiceNow. It automates the identification of gaps, parses discovery logs, determines root causes, and suggests remediation actions---all accessible through the ServiceNow Otto panel without manual log file navigation. This workflow is designed for Discovery administrators to efficiently diagnose and resolve attribute issues in the CMDB. Show full answer Show less ## Requirements and Configuration * **Applications and Plugins:** ServiceNow Otto for IT Operations Management (ITOM) must be installed, and the Discovery and Service Mapping plugin (com.snc.discovery) must be active. * **User Role:** The user must have the `discoveryadmin` role. * **AI Search:** Enabling AI Search allows natural language queries to be matched to CI tables. Without AI Search, direct table names must be provided in queries. * **Property Configuration:** The `glide.discovery.savepatternlog` property should be set to `true` to ensure pattern logs are saved for analysis. If set to `false`, the workflow may not be able to analyze missing attributes properly. ## How It Works Discovery administrators initiate the workflow via natural language questions or by selecting options in the Discovery Admin Workspace's Otto panel. The workflow supports: * **Exploratory investigation:** Surfacing available CI attributes for selection. * **Direct investigation:** Querying specific CI classes and attributes. The workflow identifies affected CIs missing the specified attribute, analyzes recent discovery logs (up to 30 days), determines root causes, displays error codes when classified, and suggests remediation actions. Two agents operate within the workflow: * **Pattern Diagnostic agent:** Processes the query and performs autonomous log analysis. * **Error Framework Remediation agent:** Provides remediation suggestions when the root cause is classified and logged in the Error Framework. If classification or logging fails, the workflow concludes with a diagnostic summary listing affected CIs, error messages, and general troubleshooting advice. ## Scope and Limitations * Supports only CIs discovered through pattern-based Discovery. * Analysis relies on default pattern metadata in the `sndiscoaipatternmetadata` table; custom pattern changes are not considered. * Examines up to five sample CIs per attribute issue. * Log analysis covers the last five days by default, extending to 30 days if no relevant data is found. ## Practical Examples * **Linux Server (Direct infrastructure):** Investigates a missing `cpumanufacturer` attribute by resolving the CI's IP address and querying discovery logs directly, returning a detailed report with root cause and remediation. * **MySQL Database (Applicative pattern):** Investigates missing attributes by navigating the "Runs on" relationship to the host server's IP address and querying host discovery logs. * **AWS Auto Scaling Group (Cloud pattern):** Investigates missing `clusterconnection` attribute by resolving the SA-LDC identifier via the "Hosted on" relationship and querying discovery logs using this identifier and process ID. The Pattern Diagnostic agentic workflow helps Discovery administrators investigate missing CI attributes. It identifies the gap, parses discovery logs, identifies the root cause, and suggests remediation --- without manually navigating log files. When Discovery runs, it populates CI attributes in the CMDB using both probe-based and pattern-based discovery. The Pattern Diagnostic agentic workflow supports investigation into CIs discovered through pattern-based discovery. When an attribute is missing, identifying the cause requires navigating multiple tables and interpreting nested JSON in discovery logs. The Pattern Diagnostic agentic workflow automates this investigation and suggests a remediation action, all from the ServiceNow Otto panel.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-1} ## Requirements {#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-requirements} ServiceNow Otto for IT Operations Management (ITOM) must be installed on your instance. For more information, see [Install the ServiceNow Otto for IT Operations Management (ITOM) application](https://www.servicenow.com/docs/hwRPYCUe1vf5jAprSqnh1Q "Install the ServiceNow Otto for IT Operations Management (ITOM) application.").{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-8} The Discovery and Service Mapping plugin (com.snc.discovery) must be active for this workflow to be available.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-13} Role required: discovery_admin.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-7} Note: AI Search must be enabled for the workflow to match natural language queries to CI tables. If AI Search is not enabled, you can still use the workflow by providing the direct table name in your query. To enable AI Search, navigate to AllAI SearchEnable AI Search. After enabling, verify that the search status is enabled to confirm indexing is complete. ## Required property configuration {#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-property} The glide.discovery.save_pattern_log property controls whether pattern logs are saved after successful pattern execution. By default, this property is set to true, which saves all pattern logs.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-9} If this property is set to false, successful pattern logs are not saved. This can prevent the Pattern Diagnostic agentic workflow from performing missing attribute analysis in cases where the attribute is not failing the pattern. Verify that this property is set to true to enable complete analysis. For information about configuring this property, see [Discovery properties](https://www.servicenow.com/docs/0irr66vIpUl6GxbWD~edYg "Discovery properties allow you to control several aspects of the horizontal discovery process.").{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-10} ## Pattern Diagnostic agentic workflow overview {#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-overview} A Discovery administrator triggers the workflow from the ServiceNow Otto panel in Discovery Admin Workspace by asking a natural-language question or by selecting the corresponding pill. The workflow supports both exploratory investigation, where it surfaces available attributes for selection, and direct investigation, where the administrator specifies the CI class and attribute in the initial question. The workflow then identifies affected CI records with the selected attribute missing, parses the relevant discovery logs to identify the root cause and, when the root cause is classified, displays the corresponding error code, and suggests a remediation.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-2} The workflow uses two agents:{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-3} * Pattern Diagnostic agent: Receives the administrator's query and runs script tools autonomously to investigate the missing attribute. * Error Framework Remediation agent: Receives the identified root cause and suggests a remediation. The second agent runs only when the root cause is classified and logged to the Error Framework. If classification or logging doesn't succeed, the workflow ends after the diagnostic summary, which lists the affected CIs, the error messages from the discovery logs, and general troubleshooting guidance.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-12} When the Error Framework plugin is active and remediations are mapped for the identified error, the workflow provides enhanced remediation suggestions.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-11} ## Scope {#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-scope} The workflow covers attribute coverage gaps for CI classes discovered through pattern-based Discovery. The following conditions apply:{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-6} * Only CIs discovered through patterns are supported for analysis. * Analysis is based on the pattern metadata table (sn_disco_ai_pattern_metadata). This table contains metadata about the patterns populating a CI and its attributes. Only default pattern information is included; customizations made to patterns aren't considered. Metadata population is limited to an agreed set of CI classes. * Log analysis focuses on the first CI in the sample. If issues are found, the workflow reports them and stops. If no issues are found, it checks the next CI in the list, up to the sample limit of five. * Log analysis covers the last five days. If no results are found in that period, the search expands to 30 days. ## Examples {#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-example} The following scenarios illustrate how the workflow handles different CI and pattern types.{#pattern-diagnostic-agentic-workflow__pattern-diagnostic-agentic-workflow-para-4} Scenario 1: Direct infrastructure (Linux Server) : The workflow investigates a missing `cpu_manufacturer` attribute on Linux Server CIs. The workflow resolves the CIs IP address and queries the discovery logs directly. It returns a full report with sample CIs investigated, link to CIs, root cause and remediation guidance. Scenario 2: Applicative pattern (MySQL database) : The workflow investigates a missing attribute on a MySQL database CI. MySQL CIs are discovered through their host servers. The workflow navigates the Runs on relationship to resolve the host server's IP address and queries the discovery logs on the host, not the MySQL instance. It returns a full report with sample CIs investigated, link to CIs, root cause and remediation guidance. Scenario 3: Cloud pattern (AWS Auto Scaling Group) : The workflow investigates a missing `cluster_connection` attribute on an AWS Auto Scaling Group CI. Cloud CIs use an SA-LDC identifier instead of a traditional IP address. The workflow navigates the Hosted on relationship to resolve the SA-LDC identifier and searches logs using the identifier and process ID. It returns a full report with sample CIs investigated, link to CIs, root cause and remediation guidance. *[\>]: and then