The Real-Time Insights Diagnostics page shows the health of the Real-Time Insights service, including its database and ingestion pipeline. You can also look up a specific endpoint and export support data to share with Nerdio Support. The page shows the Real-Time Insights version number under its heading.
-
In Nerdio Manager, navigate to Observability > Insights > Real-Time.
-
Press Ctrl+Alt+Shift+D. The Diagnostics page opens with the following tabs: System Status, Stored Procedures, Database Activity, Azure Tables, Endpoint Lookup, and Export Support Data.
The System Status tab shows when the status was last checked. Select Refresh to check again.
The Instance health tiles show the state of each Real-Time Insights component:
-
API and Worker: Whether the service is healthy, the version, and how long it has been running. The worker tile also shows how many jobs are running correctly.
-
SQL database: The database response time and the number of migrations pending.
-
Table storage: The storage response time and the number of tables that can be reached.
-
App Insights: Whether Application Insights is configured, and whether its logs are available for export.
-
Stored procedures: How many background database jobs are running correctly.
The Ingestion pipeline table shows, for each batching component, its status, any backlog, and when it last reported. A row is flagged when it has not reported within its expected cadence. The components are the app reliability event batcher, the event log batcher, the last seen batcher, the last value batcher, and the metric status batcher.
Expand All recorded diagnostic entries to see the raw diagnostic key and value pairs. If nothing has been recorded, the page says so.
The Stored Procedures tab shows the background maintenance jobs that run in the Real-Time Insights database. At the top, the runner status shows whether the job runner is running, and how many jobs there are, how many are succeeding, and how many are failing. The runner state is only written when it changes, so use the job timestamps to confirm that jobs are running.
The Jobs table lists, for each job, its status, how often it runs, its last run, its next run, its duration, and its parameters. Failing jobs are listed first, then jobs by their next run.
The Database Activity tab shows a snapshot of what the Real-Time Insights database is doing now. The tiles show how many user requests are running, how many are blocked, and the longest running request. Below them:
-
Requests in flight lists the running requests, longest first, with the session, elapsed time, status, what each is waiting on, which session is blocking it, the caller, and the statement.
-
Most expensive statements lists the statements with the highest average run time, with their average, worst, and total time, the number of runs, the last run, and the statement.
Note
After you refresh the snapshot, a countdown shows how long you must wait before you can refresh it again.
The Azure Tables tab lets you read the data that Real-Time Insights stores in Azure tables.
-
Optionally, to read the data for a single device or session user, enter a host name, endpoint ID, Intune device ID, UUID, or session user in the In focus box, and then select Find. The Azure Tables and Endpoint Lookup tabs both follow the device or user you find. If you leave nothing in focus, the tab reads the whole day.
-
From the Data list, select the data to read, for example Endpoint metrics.
-
Select the Day (UTC), Start time (UTC), and End time (UTC). The start and end times default to the whole day.
-
Select Show rows. To download the rows, select CSV.
Note
The day and times on this tab are in UTC, not your local time. The day list only holds days that are stored, so an empty result means the data is outside the retention period, not that collection failed.
The Endpoint Lookup tab shows what Real-Time Insights knows about a device, including every error that its collector has uploaded.
-
In the Device box, enter a host name, endpoint ID, or system UUID.
-
From the Errors from list, select how far back to show errors, for example Last 24 hours.
-
Select Look up.
Use the Export Support Data tab to download a diagnostics package that you can send to Nerdio Support.
-
From the Timeframe list, select the period to include. The range that results is shown below the list.
-
From the Minimum log level list, select the lowest log level to include. The options are 0 - Trace and above, 1 - Debug and above, 2 - Information and above, 3 - Warning and above (the default), 4 - Error and above, and 5 - Critical only.
-
Under Included data, clear the check box of any group of data sets you do not want to include. The Lookups and configuration group is always included.
-
Select Download diagnostics.
The following table lists the groups of data sets in the package. By default, the package contains all 43 data sets.
|
Group |
Data sets |
Contents |
|---|---|---|
|
Core data |
7 |
Endpoints, users, sessions, alerts, log events, applications, and application versions. |
|
Metrics and thresholds |
5 |
Endpoint and session metric status, last values, and raw endpoint metrics. |
|
Lookups and configuration |
8 |
Metric definitions, groups and targets, severity levels, session states, threshold configuration, polling configurations, and settings. Always included. |
|
Portal views |
6 |
The data behind the endpoints, sessions, users, alerts, host pools, and host overview views. |
|
Logon, profile and drive detail |
6 |
Session logon details, FSLogix details, endpoint volumes and daily volume data, system diagnostics, and stored procedure jobs. |
|
Application reliability |
9 |
Application reliability events, apps, hosts, users, versions, and hourly, five-minute, and daily summaries. |
|
Application Insights logs |
2 |
Traces and exceptions. |
Note
The package never includes credentials or device keys. Settings whose names look like they hold credentials are exported with their values replaced. The package lists every data set in a manifest.json file, with its row count and status, including data sets that came back empty.
Each endpoint also has its own Endpoint Diagnostics page, which includes the Polling Interval tab. For details, see Insights: Real-Time.
Comments (0 comments)