Event Center¶
Event Center is a unified event query and analysis platform for TrueWatch. It aggregates event data from multiple sources, including Monitors, integration collectors, Smart Monitoring, AI Monitors, SLO, Security Monitoring, Audit, and OpenAPI, and provides multi-dimensional filtering, detail viewing, and correlation analysis capabilities.
Event Center only performs event aggregation and querying. It does not generate events, configure rules, or send notifications. Events are generated independently by each source module and automatically aggregated here.
Event Sources¶
| Source | Generation Method | df_source Value |
|---|---|---|
| Monitor | Configure detection rules in the monitoring module and trigger when conditions are met | monitor |
| Smart Monitoring | Detect anomalies according to built-in rules or fixed rules generated by AI and saved | smartMonitor |
| AI Monitor | At each execution, AI analyzes observability data based on a natural language Prompt and generates an event report when it finds an issue supported by evidence | aiMonitor |
| Change Events | Records of runtime activity events in Kubernetes clusters that differ from expected patterns | change |
| SLO | Configure service level objectives in the SLO module and trigger on breach | slo |
| Security Monitoring | Anomalous behavior or vulnerability alerts detected by the security module | security |
| Audit Events | Records of user operations on the TrueWatch platform | audit |
| Custom Events | Actively written through OpenAPI | user |
| Integration Collection Events | Anomalies and recovery events generated during cloud collector runtime | monitor (df_monitor_type=integration_collection) |
Event Content Composition¶
Event content is determined by the source system when the event is generated. Event Center only displays it.
Rule Monitor Events¶
Monitor event content is based on the template configured in Create > Event Notification and supports dynamic filling through variables:
| Variable (Example) | Description | Example |
|---|---|---|
{{Result}} |
Detection result value | 85.5 |
{{df_dimension_tags}} |
Event dimensions | {"host":"web01"} |
{{df_monitor_checker_name}} |
Monitor name | Host CPU Monitor |
The system replaces variables based on actual monitoring data to generate the final event content.
As shown below, define the event title as Log Detection - Multiple Indexes. The event content includes DQL query statements and variables, and the system generates and displays the final result based on actual monitoring data.
After a rule detects an anomaly, you can view the related event content in Events > Event Details.
Smart Monitoring Event Analysis Report¶
Smart Monitoring executes detection according to saved fixed rules. When rules meet the trigger conditions, the system generates Smart Monitoring events and provides an analysis report in the event details.
Multiple rules may be triggered simultaneously in the same detection. The system summarizes the rules hit in this detection into a single analysis report and displays information such as anomaly summary, rule details, monitoring signals, current values, determination basis, anomaly causes, and trends by detection dimension. Smart Monitoring events have df_source set to smartMonitor; use df_source:smartMonitor to filter and view them.
AI Monitor Event Report¶
AI Monitor uses the Prompt as a continuously effective detection rule and queries and analyzes relevant observability data at each execution. After identifying issues that match the description and are supported by data evidence, the system generates a structured event report and displays it as an event record in Event Center.
The event report provides an issue summary, confidence level, and supporting evidence, and supplements it with impact scope, diagnosis, and recommended actions to help users understand the issue, confirm the impact, and carry out follow-up handling.
AI Monitor event reports have df_source set to aiMonitor. Go to Event Center > All Events and use df_source:aiMonitor to filter and view them.
Integration Collection Events¶
Associated events of cloud collectors are used to view collection anomalies and recovery status and help locate data collection issues. Different anomalies display their respective recovery status, and historical anomaly events are retained for retrospective analysis.
Go to Event Center > All Events, select a time range, and enter df_source:monitor AND df_monitor_type:integration_collection in the search bar. You can further narrow the scope by cloud provider, integration product, cloud account, and event status. For detailed steps and field examples, see Query Integration Collection Events in Event Center.
Event Field Reference¶
Common Fields¶
The final event record generally contains the following fields:
Field |
Description |
|---|---|
date / timestamp |
Generation time. Unit: seconds |
df_date_range |
Time range. Unit: seconds |
df_check_range_start |
Detection range start time. Unit: seconds |
df_check_range_end |
Detection range end time. Unit: seconds |
df_issue_start_time |
Time when the first fault in this round occurred. Unit: seconds |
df_issue_duration |
Duration of the current fault round, in seconds (from df_issue_start_time to this event) |
df_source |
Event source. For common values, see Event Sources |
df_status |
Event status. Includes ok, info, fatal, warning, error, critical, nodata, nodata_ok, nodata_as_ok, manual_ok. For the mapping between monitor event severity and status, see Event Severity |
df_sub_status |
Event detail status (supplement to df_status) |
df_event_id |
Unique event ID |
df_title |
Title |
df_message |
Description |
Fields Specific to Monitor Events¶
Rule Monitor events have df_source = monitor. The following are field references for rule monitors. Integration collection events use the same df_source and are distinguished by df_monitor_type = integration_collection. For viewing instructions, see Integration Associated Events.
Field |
Description |
|---|---|
df_dimension_tags |
Detection dimension tags, such as {"host":"web01"} |
df_monitor_id |
Alert Policy ID |
df_monitor_name |
Alert Policy name |
df_monitor_type |
Type: custom for custom monitoring events, slo for SLO events, and fixed to bot_obs for Intelligent Inspection events |
df_monitor_checker |
Execution function name, such as custom_metric, etc. |
df_monitor_checker_sub |
Detection stage: nodata when generated during data gap detection, and check when generated during normal detection |
df_monitor_checker_id |
Monitor ID |
df_monitor_checker_name |
Monitor name |
df_monitor_checker_value |
Anomaly value when the event was generated |
df_monitor_checker_value_dumps |
Anomaly value when the event was generated (JSON serialized) Makes it convenient for consumers to obtain the original value through deserialization |
df_monitor_checker_value_with_unit |
Anomaly value when the event was generated (optimal unit) |
df_monitor_checker_ref |
Monitor association; a field associated only with the DQL statement in the detection configuration |
df_monitor_checker_event_ref |
Monitor event association; a field associated only with df_dimension_tags and df_monitor_checker_id |
df_monitor_ref_key |
Association Key for self-hosted inspection, used to correspond to self-hosted inspection |
df_fault_id |
Current fault round ID; the value is the df_event_id of the first fault event |
df_fault_status |
Current fault round status; a redundant field for df_status and df_sub_status, indicating whether it is OK. Values are as follows:ok: normal fault: fault |
df_fault_start_time |
Start time of the current fault round. |
df_fault_duration |
Duration of the current fault round, in seconds (from df_issue_start_time to this event) |
df_event_detail |
Event detection details |
df_user_id |
Operator user ID during manual recovery |
df_user_name |
Operator username during manual recovery |
df_user_email |
Operator user email during manual recovery |
df_crontab_exec_mode |
Execution mode; possible values are as follows.crontab manual |
df_site_name |
Current TrueWatch site name |
df_workspace_name |
Name of the workspace to which it belongs |
df_workspace_uuid |
UUID of the workspace to which it belongs |
df_label |
Monitor labels; labels specified in the Monitor are stored in this field UUID |
df_alert_policy_ids |
Alert Policy IDs (list) |
df_alert_policy_names |
Alert Policy names (list) |
df_matched_alert_policy_rules |
Alert Policy names and all matched rule names (list) |
df_channels |
List of Incident channels to which the event belongs |
df_at_accounts |
@ account information |
df_at_accounts_nodata |
@ account information (data gap) |
df_message_at_accounts |
List of detailed @user information in fault alert messages |
df_nodata_message_at_accounts |
List of detailed @user information in data gap alert messages |
df_workspace_declaration |
Workspace Attribute Claims |
df_matched_alert_members |
When sending by member is selected, list of all matching alert notification member information |
df_matched_alert_upgrade_members |
When sending by member is selected, list of all matching alert escalation notification member information |
df_matched_alert_member_groups |
When sending by member is selected, names of all matching member groups |
df_charts |
Chart information appended when charts are added in the Monitor configuration and this alert event needs to send a message |
df_alert_info |
Records alert notification information |
df_is_silent |
Whether the event is muted; the value is the string "true" / "false" |
df_sent_target_types |
List of non-duplicate alert notification target types that have been sent for this event |
df_check_targets |
Records the detection time and detection metrics in the Monitor configuration |
df_check_condition, df_check_condition_expr |
Records the trigger rules in the Monitor configuration |
df_check_rules |
Records the detection count and detection rules in the Monitor configuration |
df_check_targets |
Records the detection time and detection metrics in the Monitor configuration |
df_check_condition |
Records the trigger rules in the Monitor configuration |
df_check_interval |
Records the detection frequency in the Monitor configuration |
df_fault_end_time |
Records the fault recovery time |
Fields Specific to Smart Monitoring Events¶
When df_source = smartMonitor, df_event_report records Smart Monitoring analysis report data. Multiple rules hit in the same detection are summarized in this analysis report.
Fields Specific to AI Monitor Events¶
When df_source = aiMonitor, you can use the following fields to identify the event source and view the data on which AI analysis is based:
Field |
Description |
|---|---|
df_monitor_type / df_monitor_checker_type |
Fixed to ai, indicating that the event was generated by an AI Monitor. |
df_monitor_checker_id / df_monitor_checker_name |
ID and name of the AI Monitor that generated the event. |
df_message |
Structured event report generated by AI. |
df_check_targets |
Data objects and query information used in this detection, used to correlate analysis evidence. |
Fields Specific to Audit Events¶
When df_source = audit, the following additional fields exist:
| Field | Description |
|---|---|
df_user_id |
Operator user ID |
df_user_name |
Operator username |
df_user_email |
Operator user email |
| {Other fields} | Other fields based on specific audit data requirements |
Fields Specific to Custom Events¶
When df_source = user, the following additional fields exist:
| Field | Description |
|---|---|
df_user_id |
Creator user ID |
df_user_name |
Creator username |
df_user_email |
Creator user email |
| {Other fields} | Other fields generated based on user operations |
Event Query and Viewing¶
Event Center provides multiple Explorers to meet event query and analysis needs in different scenarios.
| Event View Dimension | Description | Core Capabilities |
|---|---|---|
| All Events | Full event query | Aggregates and displays event data from all sources in the current workspace, supports both list and chart display modes, and provides multi-dimensional filtering, search, export, snapshot saving, and more |
| Unrecovered Events | Active fault tracking | Concentrates events with df_status != ok in the last 48 hours and displays fault duration aggregated by df_fault_id |
| Change Events | K8s change audit | Records runtime activity events in Kubernetes clusters that differ from expected patterns and displays configuration differences before and after changes |
| Smart Monitoring Events | Smart anomaly viewing | View anomaly events detected by Smart Monitoring rules and analysis reports generated by merging multiple hit rules in a single detection |
| Event Details | In-depth analysis of a single event | Click any event to enter and view complete information such as basic attributes, extended fields, alert notifications, historical records, related events, and related SLOs |
Further Reading¶
After understanding the basic concept of "event", you may need:

