Documentation Index

Fetch the complete documentation index at: https://kb.vastdata.com/llms.txt

Use this file to discover all available pages before exploring further.

Viewing Logs and Traces

Prev Next

Viewing Pipeline Logs

Logs are collected using the Open Telemetry framework.

To view logs:

  • To see all logs, open the Pipeline Management page from the left navigation menu and select Logs.

    To see logs for a specific pipeline, right-click the pipeline and select Pipeline Logs. This opens the Logs page and filters the logs for the pipeline you selected.

    The Logs page displays the following properties for each log:

    Timestamp

    The date and time of the logged event.

    Level

    The severity level:

    • TRACE. A fine-grained debugging event.

    • DEBUG. A debugging event.

    • INFO. An informational event. Indicates that an event happened.

    • WARN. A warning event. Not an error but is likely more important than an informational event.

    • ERROR. An error event.

    • FATAL. A fatal error such as application or system crash.

    Scope

    The scope that emitted the log:

    • user. The log was emitted by the function code.

    • vast-runtime. The log was emitted by the VAST DataEngine runtime.

    Pipeline

    The pipeline that emitted the log.

    Message

    A description of the logged event.

You can also inspect a log:

Select a log in the Logs page and click the Inspect button to open the Inspect panel. The following details are shown:

  • The log message, if it is long.

  • The resource.

  • A Trace ID, with a hyperlink to the expanded trace display.

  • A span ID associated with the log.

  • A mini version of the trace.

Viewing Pipeline Traces

Traces are collected using the OpenTelemetry framework and observatory system, which enables you to see how each trigger event traverses through the pipeline, how time is distributed between function deployments, how traces are associated with specific attributes and logs, which function deployment events are invoked by a given trigger event, and provides visibility into performance bottlenecks.

You can view a list of traces and you can explore each trace for more visibility.

Viewing All Traces

To see all traces, open the Pipeline Management page from the left navigation menu and select Traces.

To see traces for a specific pipeline, right-click the pipeline and select Pipeline Traces. This opens the Traces page and filters the traces for the pipeline you selected.

The following properties are shown per trace:

Start Time

The start time of the trace.

Pipeline

The pipeline that emitted the trace.

Status

The status of the trace.

Trace ID

The trace ID. Click the trace ID to explore the trace.

Duration

The duration of the trace.

Exploring a Trace

When you click a trace ID, the user interface shows you one or more spans of the trace in the pipeline's runtime, as color coded bars.

To see logs associated with the trace, click Logs.

To see logs associated with a component span, click the relevant bar.

The following attributes are also displayed for the trace:

  • Trace Start Time

  • Duration

  • Functions

  • Depth

  • Total Span

Batch Traces

When a function processes events in batches, there is a trace for each event and for the batch.

On the Traces page, you can select Batch Traces from the View toggle to list batch traces.

For batch traces, the number of events is shown per trace.

Traces for routed functions

When you select a trace ID of a trace from a routed function, you can select one of the following ways to view the trace from the View toggle: Flame or Waterfall.

The Waterfall view displays every span executed within the trace, as bars stacked in chronological order. Each bar's length indicates how long that specific step took, and its indentation shows parent-child relationships between functions.

Hovering over the trace bar in the Waterfall view brings up a pipeline diagram on the right. The pipeline diagram highlights the path that the traced event took during the run.