Skip to main content

Task errors: Studio views

  • Tasks > Task Errors
    Open the Task Errors view from the Tasks menu to inspect errors raised by ETL and AI tasks.
    You can browse all errors in a unified list or group them by task, apply various filters, select an error to view it in detail, and see how task health is impacted by recent errors.

  • AI Hub > AI Task Errors
    The AI Task Errors view, opened from the AI Hub, is a pre-filtered subset of the Task Errors view.
    Use this view to inspect errors raised by Embeddings Generation and GenAI tasks.

  • Both views display the same errors for listed tasks; deleting a task's errors from one view is reflected in the other.

  • Tasks > Ongoing Tasks
    Each ETL and AI task bar on the Ongoing Tasks view shows the task's health state and error count; expanding the bar reveals additional detail.

  • To learn about task errors and how they impact task health, see the Overview page.

  • In this article:

Task Errors view​

In its default layout, Task Errors groups errors into per-task segments, each showing the task's errors in a sortable table.

Opening the view​

Open the Task Errors view from the Tasks menu. By default it will open with no filters applied, showing a segment for every ETL or AI task that currently has any errors.

Task Errors view

  • A. Click to open the Tasks menu.

  • B. Click to open the Task Errors view.

  • C. Task filters (see below).

  • D. Task health indicators (see below).

  • E. Toggle to group errors by task or display them in a unified list.

  • F. Task errors (see below).

Task filters​

Use the filters bar to narrow the listing to specific tasks and errors.

Filters bar

  • Filter by task/script name
    Type a task or script name to narrow the listing to matching tasks.

  • Filter by node
    Pick one or more cluster nodes to show only the errors raised on the selected nodes.

  • Filter by task type
    Pick one or more task types (e.g., Kafka ETL) to show only the errors raised by the selected types.

  • Filter by task health
    Pick one or more health states to show only tasks currently in the selected states.

Task health indicators​

The indicators' colors represent task health states: Green for Healthy, yellow for Impaired, and red for Failed.

  • Hover an indicator to trigger a popup summary of tasks whose health currently matches the selected state.
  • The summary lists only the node currently running the task and any nodes that recorded errors for it, with the error count per node.

Task health indicators

Task errors​

The image below shows one of the task segments displayed in the task errors view when errors are grouped by task.

Task Errors - per-task segment

  1. Task name
    The name of the ETL or AI task whose errors are displayed here.

  2. Delete errors
    Click to remove all errors raised by this task, including both item and process errors.

    Deleting a task's errors does not, on its own, reset the task's health state.
    Health is driven by the running error rate, not by the rows in the error tables.
    A task in Impaired or Failed state will recover only as new batches complete successfully and its error rate falls back below the configured thresholds.
    See the Overview for more.

  3. Task metadata row

    • A toggle to collapse or expand all errors related to this task.
    • Task type.
    • Error count for this task.
    • The number of scripts that this task runs.
    • Task's current health state (Healthy, Impaired, or Failed).
    • Tag/s of the cluster node/s currently running the task.

  4. Script sub-segment details
    Errors for each script the task runs appear in their own sub-segment, with a header showing the script's name and error count and a toggle to collapse or expand the errors related to this script.

  5. Errors table
    The script's errors, one row per error.

    • Column headers
      You can filter or sort the table by the content of each column, using the funnel (filter) or arrow (sort) icons at the column header.

    • Show column
      Click the eye icon for a specific error to open an error-details dialog with the full error message.

    • Error type column
      Marks the row as Item Error (a single document failure the task skipped past) or Process Error (a batch-scope failure that may affect multiple documents).

    • Error step column
      The processing step the error occurred at: Configuration, Extraction, Transformation, Load, Persistence, Model Inference, or Unknown.
      See the Error steps reference on the overview.

    • Document column
      For item errors, the ID of the document being processed when the error occurred, rendered as a hyperlink to the document.
      For process errors, the column shows - because the error is not bound to a single document.

    • Date column
      The error's creation timestamp, shown in date form and in relation to the current time (e.g., "4 hours ago").

    • Affected Documents column
      For process errors, the number of documents the failing batch attempted to process.
      Empty for item errors.

    • Error column
      The error message, truncated to one line.

    • Node column
      The tag of the cluster node that recorded the error.

AI Task Errors view​

The AI Task Errors view lists the same errors listed by the Task Errors view, with the same layout, controls, and data, but applies a predefined filter to show only Embeddings Generation and GenAI task errors.

All options documented under Task Errors view above apply here without change.

AI Task Errors view

  1. Click to open the AI Hub.

  2. Click to open the AI Task Errors view.

Task health on the Ongoing Tasks view​

On the Ongoing Tasks view, each ETL or AI task bar displays the task's current health state and the number of errors recorded for the task. Expanding the bar reveals these details per node, along with the node's Connection status.

Collapsed view​

Ongoing Tasks view

  1. Click to open the Tasks menu.

  2. Click to open the Ongoing Tasks view.

  3. Task bar

  4. Expand details
    Click to expand the bar - see the per-node breakdown below.

  5. Task health and error count

    • Health Status - the task's current state (Healthy, Impaired, or Failed).
    • Errors - the number of errors currently recorded for the task.

Expanded view​

Ongoing Tasks task bar - expanded

Each relevant node has its own column showing how the task is doing on that node. Only the node currently running the task, and any other nodes that recorded errors for it, are shown.

  • Connection status - the state of the node's connection to the task's destination. The value is Active while the connection is up, and Reconnect after a failure while the task waits to retry.
    To retry the failing batch immediately, hover Reconnect and click the Retry now button that appears.

  • Errors - the number of errors the task has raised on this node.

  • Health status - the task's classification on this node (Healthy, Impaired, or Failed).

  • State - the task's processing state on this node (such as UP TO DATE or 0% RUNNING).


See the Ongoing Tasks - Overview page for a full walkthrough of the view, including filters, selection, and per-task actions.

In this article