Monitors
Types of Monitors
A Monitor is a core component of the monitoring and alerting system. Its main responsibility is to observe events, states, or conditions within the environment and determine whether a specific situation has occurred.
Monitors can be configured to track a variety of aspects, such as:
- The state of connectors (e.g., running, failed, stopped).
- The health of Kafka clusters.
- Changes in system metrics.
- Scheduled conditions, either within defined time intervals or at exact points in time.
When a monitor detects that its condition is met, it triggers an associated Action. This flexible approach allows users to define what is important to observe and decide how the system should respond, ensuring proactive management and automation.
All available types of Monitors are presented in the table below.
Icon | Display Name | Name | Description |
|---|---|---|---|
 | Connector Health | CONNECTOR_STATE | Monitor Kafka Connect connectors for status changes |
 | Kafka Health | KAFKA_STATE_SCHEDULED_TASK | Monitor Kafka cluster health and availability |
 | Scheduled Timepoint | TIME_RANGE_SCHEDULED_TASK | Run actions at specific intervals |
 | Interval Action Runner | DELAYED_INSTANT_TASK | Run actions at scheduled time |
 | Metric Condition | CONNECTOR_METRIC_CHANGE_SCHEDULED_TASK | Trigger actions based on specific metric changes |
Deleting a connector linked to a monitor does not remove the monitor itself. The monitor remains stored in the system configuration but may not function correctly until a working connector is reassigned.
Connector Health
The Connector Health Monitor is designed to track the state of a selected connector within the system. Each connector can exist in one of several states:
- RUNNING – the connector is active and functioning normally.
- RESTARTING – the connector is in the process of restarting.
- FAILED – the connector has encountered an error and stopped working.
- PAUSED – the connector is temporarily suspended.
- STOPPED – the connector has been deliberately stopped.
- UNASSIGNED – the connector is not currently assigned to any worker.
The monitor reacts only to specific connector state changes. It tracks situations where the connector transitions to FAILED or, during a restart, moves to RUNNING. When such a change is detected, the monitor can automatically trigger the assigned action, such as sending a notification or attempting to restart the connector.
This monitor allows users to respond only to critical state changes, ensuring rapid reaction and maintaining the stability of data processing pipelines.
When creating a Connector Health Monitor, several fields must be defined to describe its behavior. These include:
| Field Name | Requried | Description |
|---|---|---|
Name | Yes | Is used to recognize and display the monitor within the system, allowing users to easily distinguish it from other monitors. |
Description | No | An optional text field where the user can provide additional information about the monitor. |
Connector Name | Yes | Specifies the connector that the monitor will observe. This parameter defines the exact connector whose state changes will be tracked, allowing the monitor to react appropriately when the connector’s status changes. |
Interval | Yes | The time period at which the monitor checks the state of the selected connector. By default, this value is set to 3600 seconds. Adjusting the interval allows users to control how frequently the system verifies connector status and triggers actions if needed. |
Kafka Health
The Kafka Health Monitor is designed to observe the overall state of a Kafka cluster. The cluster can exist in one of the following states:
- UP – the Kafka cluster is operational and functioning normally.
- DOWN – the Kafka cluster is unavailable or has encountered a failure.
When the monitor detects a state change, it can automatically trigger the associated Action. For example, if the cluster goes DOWN, the monitor might send a notification or initiate a recovery procedure. This monitor enables users to proactively track the health of the Kafka infrastructure and respond immediately to issues, ensuring data pipelines remain stable and reliable
When creating a Kafka Health Monitor, several fields must be defined to describe its behavior. These include:
| Field Name | Requried | Description |
|---|---|---|
Name | Yes | Is used to recognize and display the monitor within the system, allowing users to easily distinguish it from other monitors. |
Description | No | An optional text field where the user can provide additional information about the monitor. |
Interval | Yes | The time period at which the monitor checks the state of the Kafka cluster. By default, this value is set to 3600 seconds. Adjusting the interval allows users to control how frequently the system verifies cluster status and triggers actions if needed. |
Interval Action Runner
The Interval Action Runner Monitor is designed to execute assigned actions within a specified time period. The time range can be either finite or indefinite. During the selected period, the monitor automatically runs the assigned actions.
This type of monitor allows users to restrict action execution to specific hours or days, ensuring that operations occur only within the planned time window.
When creating a Interval Action Runner Monitor, several fields must be defined to describe its behavior. These include:
| Field Name | Requried | Description |
|---|---|---|
Name | Yes | Is used to recognize and display the monitor within the system, allowing users to easily distinguish it from other monitors. |
Description | No | An optional text field where the user can provide additional information about the monitor. |
Start Time | Yes | Specifies the exact time when the monitor becomes active. The monitor will not operate before this time, and it is not possible to set a time in the past. This ensures that monitoring and associated actions only begin at the intended moment. |
End Time | Yes | Specifies when the monitor should stop being active. Users can set a specific time in the future or leave it indefinite, allowing the monitor to continue running without a fixed end. This controls the duration of monitoring and ensures actions are only executed within the desired time range. |
Interval | Yes | Specifies the time interval at which the assigned action will be executed within the defined time range. This parameter determines how frequently the action runs during the monitor’s active period. |
Scheduled Timepoint
The Scheduled Timepoint Monitor is designed to trigger actions at a specific, predefined time. Unlike other monitors that react to state changes or metrics, this monitor operates purely based on a schedule. For example, a Scheduled Timepoint Monitor can be used to automatically restart a connector at a particular time, or execute any other predefined action. By using this monitor, users can automate time-based tasks, ensuring that actions occur exactly when planned without manual intervention.
| Field Name | Requried | Description |
|---|---|---|
Name | Yes | Is used to recognize and display the monitor within the system, allowing users to easily distinguish it from other monitors. |
Description | No | An optional text field where the user can provide additional information about the monitor. |
Due Date Time | Yes | Specifies the exact time when the monitor triggers an action active. |
Metric Condition
The Metric Condition Monitor is designed to observe specific metrics of a selected connector. Users choose the connector they want to monitor, the particular metric, and the target value.
The monitor can be configured to react based on how the metric changes relative to a specified value, such as increasing, decreasing, or reaching a specific threshold. When the specified condition is met, the monitor triggers the assigned Action.
This monitor allows users to proactively respond to changes in connector performance or behavior, ensuring that any significant deviations in metrics are automatically detected and acted upon.
When creating a Metric Condition, several fields must be defined to describe its behavior. These include:
| Field Name | Requried | Description |
|---|---|---|
Name | Yes | Is used to recognize and display the monitor within the system, allowing users to easily distinguish it from other monitors. |
Description | No | An optional text field where the user can provide additional information about the monitor. |
Connector Name | Yes | The name of the connector to which the collected measurements refer. |
Metric Name | Yes | Specifies the particular metric of the selected connector that the monitor will observe. This field determines which aspect of the connector’s performance or behavior will be tracked, allowing the monitor to detect significant changes and trigger the assigned action accordingly. |
Options | Yes | A list of labels and filters specifying which metric values should be monitored. |
Operator Type | Yes | Defines the condition used to evaluate the selected metric against the specified value. It determines when the monitor should trigger an action based on whether the metric is LESS, LESS_OR_EQUAL, GREATER, GREATER_OR_EQUAL, EQUAL, or NOT_EQUAL to the target value. |
Target Value | Yes | Numerical value used as the reference for evaluating the selected metric. The monitor compares the metric against this value using the specified Operator Type to determine whether the assigned action should be triggered. |
Interval | Yes | The time period, in seconds, at which the monitor checks the selected metric. This determines how frequently the system evaluates the condition and triggers the assigned action if the threshold is met. |
States of Monitor
Each monitor in the system can exist in one of the following states:
- CREATED – the monitor has been created but is not yet active. Newly created monitors start in this state. To activate the monitor, it must be manually started.
- RUNNING – the monitor is active and observing the system according to its configuration. It can trigger assigned actions when its conditions are met.
- FINISHED – the monitor has completed its task and is no longer active. This state typically applies to monitors with a defined end condition or schedule.
- FAILED – the monitor encountered an error that prevented it from functioning correctly. Troubleshooting is required to resolve the issue and reactivate the monitor.
The lifecycle of a monitor begins in the CREATED state. Only after manual activation does it transition to RUNNING, allowing it to perform monitoring and trigger actions. Understanding these states helps users manage monitors effectively and respond appropriately to system events.