Health Monitoring App User Guide (v2.1.x)
Monitor integration health and troubleshoot FLA and Empower ingestion latency by using the Health Monitoring App v2.1.x dashboards.
This guide explains how to use Health Monitoring App versions 2.1.x to monitor integration health and troubleshoot the latency of one or more Tetra File-Log Agent (FLA) files or an individual Empower injection.
What's New in v2.1.0
The FLA and Empower tabs add summary tables, journey charts, and detail tables for latency troubleshooting. On the FLA tab, select one or more RAW or IDS file IDs in File Id to inspect and compare source-file journeys. On the Empower tab, select one RAW or IDS file ID to inspect a single injection. An IDS file ID resolves to the same source-file journey as its associated RAW file ID. When multiple FLA files are selected, the journey chart shows the mean of the available durations for each stage.
The existing integration-health metrics and Operational Insights workflows remain available.
Prerequisites
Before you begin, ensure you have the following:
- TDP v4.5.0 or later
- Health Monitoring App v2.1.0 or later in the v2.1.x release line, activated in your TDP environment
- A role with Health Monitoring Dashboard view access
- Agent and platform events for the files or injections you want to investigate
For activation, contact your customer account leader. For customer-hosted environments, see the Databricks, DNS, and TLS prerequisites. To change the installed app version, see Upgrade the Health Monitoring App.
Access the Health Monitoring App Dashboard
To open the dashboard, do the following:
- Sign in to the TDP as a user with Health Monitoring Dashboard view access.
- In the left navigation menu, choose Operational Intelligence. Then, select Health Monitoring.
- Select the tab for the integration you want to investigate.
Use the Dashboard Tabs
The dashboard has five tabs:
| Tab | What you can monitor |
|---|---|
| FLA | Tetra File-Log Agent scans, file uploads, ingestion failures, and file latency |
| Empower | Tetra Empower Agent projects, injection generation and upload, and injection latency |
| Chromeleon | Tetra Chromeleon Agent scans, report generation, uploads, and latency |
| Pipelines | Workflow failures, run durations, and workflow status |
| Agent Connectivity | Agent heartbeats and connection status |
The existing metrics remain available in v2.1.x. For their definitions and filters, see the FLA, Empower, Chromeleon, Pipelines, and Agent Connectivity reference sections in the v2.0.x guide. The procedures below describe the new v2.1.x latency views.
Troubleshoot File or Injection Latency
Choose a Latency View
| Purpose | FLA tab | Empower tab |
|---|---|---|
| Compare latency measurements for individual files or injections | File Latency Summary | Injection Latency Summary |
| Compare stage durations and identify the slowest stage | File Latency Journey Average Duration by Stage | Injection Latency Journey |
| Inspect exact stage timestamps, durations, and workflow status | File Latency Detail | Injection Latency Detail |
The summary tables show the file path, RAW and IDS file IDs, Agent name and ID, workflow status, and pipeline information. The Injection Latency Summary also includes project_path.
Read the Summary Latency Columns
All latency measurements are in seconds.
| Column | Measurement |
|---|---|
| latency_agent_scan_sec | Time from the file or injection change timestamp to the Agent scan timestamp |
| latency_agent_generation_sec | Time from the Agent scan timestamp to upload completion, including generation and upload time |
| latency_agent_sec | Time from the change timestamp to upload completion |
| latency_raw2ids_sec | Time from RAW file registration to IDS file registration |
| latency_E2E | Time from the change timestamp to IDS file registration |
An unavailable value means that one or more required events or timestamps are missing. Do not interpret it as zero latency.
Inspect One or More FLA Files
To inspect one file or compare multiple files, do the following:
- Open the FLA tab.
- Set Time range to include the scan event for every file you want to inspect.
- If needed, narrow the results with FLA Agent Name or FLA Agent Id.
- In File Id, search for and select one or more RAW file IDs or their associated IDS file IDs. You can mix RAW and IDS IDs.
- Review File Latency Summary to compare per-file latency measurements and identifiers.
- Review File Latency Journey Average Duration by Stage. When one source-file journey is selected, the chart shows that file's stage durations. When multiple source-file journeys are selected, the chart shows the mean of the available durations for each stage.
- Review File Latency Detail. Use fileid to distinguish each file's timestamps, durations, and workflow status.
Each selected RAW or IDS file ID resolves to its source-file journey. Selecting both IDs from the same source-file journey does not duplicate the file. The File Id filter applies to File Status and the three FLA latency views. For File Status, either a RAW or an IDS file ID resolves to the corresponding RAW file. Other FLA charts are unaffected by this filter.
Inspect a Single Empower Injection
To inspect one injection, do the following:
- Open the Empower tab.
- Set Time range to include the injection's upload-completion event.
- If needed, narrow the results with Empower Agent Name, Empower Agent Id, or Project Path.
- Enter or select a single RAW file ID or its associated IDS file ID in File Id.
- Review Injection Latency Summary, including project_path, to confirm the injection and compare its latency measurements.
- Review Injection Latency Journey and Injection Latency Detail to locate the delay and inspect its timestamps.
The File Id filter applies to these three Empower latency views. Other Empower metrics can still show results for multiple injections matching their own filters.
Interpret the Seven Journey Stages
| Stage | What the duration represents |
|---|---|
| Changed to scanned | Time between the Agent-reported change and scan timestamps |
| Scanned to upload start | Time between the scan timestamp and the start of upload |
| Upload start to upload done | Upload duration |
| Upload done to RAW registered | Time between Agent upload completion and platform RAW registration |
| RAW registered to pipeline start | Time between RAW registration and the start of the associated pipeline |
| Pipeline start to pipeline end | Pipeline execution duration |
| Pipeline end to IDS registered | Time between pipeline completion and IDS registration |
On the FLA tab, selecting one source-file journey shows that file's available stage durations. Selecting multiple source-file journeys shows the mean of the available durations for each stage across the selected files. A file with a missing duration for a stage does not contribute to that stage's mean. With File Id cleared, File Latency Journey Average Duration by Stage shows the mean for each stage across the files matching the other filters.
On the Empower tab, File Id remains single-select. Selecting one injection shows that injection's available stage durations. With File Id cleared, Injection Latency Journey shows the mean for each stage across the injections matching the other filters.
Read the Detail Table
The detail tables use the following columns. The fileid column appears only in File Latency Detail.
| Column | Meaning |
|---|---|
| fileid | File ID recorded in the FLA scan event that identifies the source-file journey (FLA only) |
| stage_label | Stage number and name |
| stage_start | Stage start timestamp |
| stage_end | Stage end timestamp |
| stage_duration_sec | Duration between the start and end timestamps, in seconds |
| stage_clock | Whether the stage uses Agent timestamps, platform timestamps, or a mixture of the two |
| workflow_status | Available status of the associated pipeline workflow |
Durations that combine Agent and platform timestamps can be negative because of clock differences. For example, the Upload done to RAW registered stage has a stage_clock value of mixed. End-to-end latency also spans Agent and platform clocks.
Limitations
- Missing events can leave stage timestamps and durations unavailable. A failed or pending conversion can have no IDS file ID or RAW-to-IDS latency.
- The FLA latency views can include files that the Agent scanned but did not upload. These files show unavailable values for upload and later stages.
- Pipeline timing and status use the earliest available workflow for the file's trace that succeeded, failed, or was canceled. Reprocessed files can therefore show an earlier run's timing and status, not the latest run. This view is not a complete pipeline-run history.
- Time range selects FLA journeys by the scan event and Empower journeys by the upload-completion event. A file outside that range can have no matching latency results even if its ID is valid.
- Each latency summary and detail table returns at most 1,000 rows. An FLA file can contribute seven rows to File Latency Detail, so large multi-file selections can reach the limit. Select fewer file IDs or narrow Time range, the Agent filters, or Project Path on the Empower tab.
- The latency views use available event records, not a live progress indicator. Missing results do not by themselves prove that a file or pipeline failed.
- The Health Monitoring dashboard does not backfill metrics when activated. Some existing charts can remain empty until enough event data is available.
The existing Health Monitoring dashboard limitations still apply to the unchanged metrics. The File Ingestion Journey chart described there is distinct from the new File Latency Journey Average Duration by Stage chart.
Troubleshooting
Missing Latency Results for Selected File IDs
- Confirm that you selected the correct tab: FLA for File-Log Agent files or Empower for Empower injections.
- Check the Time range. On the FLA tab, it must include the scan event for every selected file. On the Empower tab, it must include the selected injection's upload-completion event. A later IDS registration alone is not sufficient.
- Clear conflicting Agent filters or, on the Empower tab, Project Path, and try again.
- Confirm that every ID selected in File Id belongs to an intended source file or its associated IDS file.
- If a large FLA selection is missing detail rows, select fewer file IDs. File Latency Detail returns at most 1,000 rows.
If no results appear, contact your customer account leader or submit a support ticket with all selected file IDs, the Agent, the applicable project or path, and the selected time range.
Missing IDS or Pipeline Durations
A file can be scanned or uploaded without a completed IDS conversion. Inspect the available workflow_status and stage timestamps. Check the file's processing status before treating an unavailable duration as a failure or a zero-second conversion.
For processing troubleshooting, see Monitor Pipeline File Processing.
Negative Latency Duration
Check which timestamps the measurement uses. For stages marked mixed in stage_clock, or end-to-end measurements spanning Agent and platform events, clock differences can explain a negative value. Contact your administrator or customer account leader to investigate clock synchronization if this affects your analysis.
Dashboard Loading or Performance Problems
Refresh the dashboard and confirm your access permissions and installed app version. If queries are slow, reduce the time range and narrow the Agent filters. For the existing dashboard workflows, see Health Monitoring troubleshooting.
Related Resources
Updated about 22 hours ago

