Performance Dashboard
Overview
The Performance Dashboard is the page the application opens on once at least one server has been connected. Before that, the Welcome Page is shown instead.
It does two things:
- Lists every connected instance, one row each, and says which of them are not answering.
- Runs six reports across all of them at once, from the toolbar along the top.
That second half is the reason to come back to this page rather than treat it as a splash screen. Every other report in the product answers about one instance. These six answer about the estate, in one grid, sorted together.
Where to find it
| Route | How |
|---|---|
| Startup | Shown automatically once an instance has been connected |
| Server tree | Click the Server Health node at the top of the tree |
| Toolbar | The Dashboard button, from any of the all-instance reports |
The page title reads Database Health Monitor – Performance Dashboard.
Reading the page
The instance rows
One row per instance in your settings file, sorted by name. Each row names the instance, and clicking the name selects it in the tree, which opens its Server Overview.
An instance that is not answering says so in its row rather than being left out:
| Row reads | What it means |
|---|---|
| <name> | Connected |
| <name> (connecting…) | The connection is still being made |
| <name> (Disconnected) | It was connected and is not now, or it never answered |
The rows are built one at a time, and the progress bar underneath tracks how far through the list it is. On an estate of forty instances the page is usable long before the last row arrives, which is the point of building it that way rather than showing nothing until the slowest server answers.
Expanding an instance’s subtree is started in the background as its row is built, so the tree is ready by the time you click.
The cards above the rows
Two cards can appear above the first instance row.
| Card | When |
|---|---|
| First scan | Once, in the session after a server is first connected. It is about that server |
| Getting started | Until the checklist behind it is finished. It is about the product |
The first scan card sits above the checklist because it is the shorter lived of the two.
The all-instance reports
Six buttons along the top, with a label beside them reading <— These reports run against all connected SQL Instances.
| Button | What it opens | Offline instances |
|---|---|---|
| Dashboard | Back to this page | n/a |
| Backup Status | Backup Status – All Instances | Skipped |
| CheckDB Status | CheckDB Status – All Instances | Skipped |
| File Utilization | File Utilization – All Instances | Skipped |
| Inventory | Inventory – All Instances | Skipped |
| Failed Jobs | Failed Jobs Last 24 Hours – All Instances | Included |
| Job History | Job Schedules – All Instances | Skipped |
Each report is the ordinary single instance report, run once per instance, with every instance’s rows appended to the one grid the first instance created. The grid sorts and exports the same way any other grid in the product does, so sorting by last backup across forty instances is one click.
Failed Jobs deliberately runs against every instance, online or not. The others check that a server is answering first and skip it if it is not. A server that has stopped answering is exactly the one whose failed jobs you want to see, so that report asks anyway and reports what it finds.
What needs a license
Inventory, Failed Jobs and Job History need a registration code. Without one the button says so rather than opening an empty report:
You must enter a registration code to be able to use this feature.
Backup Status, CheckDB Status and File Utilization run on any installation.
Refresh
The page is built once, when it is opened. There is no timer behind it: the rows are a directory of what is connected rather than a live reading, and the estate does not change while you are looking at it.
Dashboard, on the toolbar, rebuilds the page. So does clicking Server Health in the tree, and F5.
Messages you may see
No servers connected. With nothing in the settings file the page is replaced by the Welcome Page, which offers three ways to connect.
You must enter a registration code to be able to use this feature. One of the three licensed buttons above was pressed on an unregistered installation.
Related reports
| Report | Why you would go there |
|---|---|
| Server Overview | Everything about one instance, rather than one row about all of them |
| Backup Status | The single instance version, with its chart |
| Inventory | Version, edition and build for one instance |
| Quick Scan Report | A scored health check, one instance at a time |
| Migration Planner | What a replacement server for one of these would need |
Frequently asked questions
Why is an instance missing from the list? Only instances in your settings file are listed. Connecting to a server adds it; removing it from the tree takes it out.
Why does a report show fewer instances than the dashboard lists? Because five of the six skip instances that are not answering. Failed Jobs is the exception and includes them.
Can I sort the combined grid by instance? Yes. It is an ordinary grid with every instance’s rows in it, so any column sorts across the whole estate.
Why does the page take a moment to fill in? It builds one instance row at a time so the window stays responsive. The progress bar tracks it.
Is this the page the tour calls the dashboard? Yes. The Server Health node at the top of the tree and the Dashboard button both return here.