diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-27 11:08:07 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-27 11:08:07 +0000 |
commit | c69cb8cc094cc916adbc516b09e944cd3d137c01 (patch) | |
tree | f2878ec41fb6d0e3613906c6722fc02b934eeb80 /web/README.md | |
parent | Initial commit. (diff) | |
download | netdata-c69cb8cc094cc916adbc516b09e944cd3d137c01.tar.xz netdata-c69cb8cc094cc916adbc516b09e944cd3d137c01.zip |
Adding upstream version 1.29.3.upstream/1.29.3upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r-- | web/README.md | 236 |
1 files changed, 236 insertions, 0 deletions
diff --git a/web/README.md b/web/README.md new file mode 100644 index 0000000..fc1d371 --- /dev/null +++ b/web/README.md @@ -0,0 +1,236 @@ +<!-- +title: "Dashboards" +description: "Every Netdata Agent comes bundled with hundreds of interactive, customizable charts designed by monitoring and troubleshooting experts." +custom_edit_url: https://github.com/netdata/netdata/edit/master/web/README.md +--> + +# Dashboards + +Because Netdata is a health monitoring and _performance troubleshooting_ system, +we put a lot of emphasis on real-time, meaningful, and context-aware charts. + +We bundle Netdata with a dashboard and hundreds of charts, designed by both our +team and the community, but you can also customize them yourself. + +There are two primary ways to view Netdata's dashboards: + +1. The [local Agent dashboard](/web/gui/README.md) that comes pre-configured with every Netdata installation. You can + see it at `http://NODE:19999`, replacing `NODE` with `localhost`, the hostname of your node, or its IP address. You + can customize the contents and colors of the standard dashboard [using + JavaScript](/web/gui/README.md#customizing-the-standard-dashboard). + +2. The [`dashboard.js` JavaScript library](#dashboardjs), which helps you + [customize the standard dashboards](/web/gui/README.md#customizing-the-standard-dashboard) + using JavaScript, or create entirely new [custom dashboards](/web/gui/custom/README.md) or + [Atlassian Confluence dashboards](/web/gui/confluence/README.md). + +You can also view all the data Netdata collects through the [REST API v1](/web/api/). + +No matter where you use Netdata's charts, you'll want to know how to [use](#using-charts) them. You'll also want to +understand how Netdata defines [charts](#charts), [dimensions](#dimensions), [families](#families), and +[contexts](#contexts). + +## Using charts + +Netdata's charts are far from static. They are interactive, real-time, and work +with your mouse, touchpad, or touchscreen! + +Hover over any chart to temporarily pause it and see the exact values presented +as different [dimensions](#dimensions). Click or tap stop the chart from automatically updating with new metrics, thereby locking it to a single timeframe. + +![Animated GIF of hovering over a chart to see +values](https://user-images.githubusercontent.com/1153921/62968279-9227dd00-bdbf-11e9-9112-1d21444d0f31.gif) + +You can change how charts show their metrics by zooming in or out, moving +forward or backward in time, or selecting a specific timeframe for more in-depth +analysis. + +Whenever you use a chart in this way, Netdata synchronizes all the other charts +to match it. + +You can change how charts show their metrics in a few different ways, each of +which have a few methods: + +| Manipulation | Method #1 | Method #2 | Method #3 | +|--- |--- |--- |--- | +| **Reset** charts to default auto-refreshing state | `double click` | `double tap` (touchpad/touchscreen) | | +| **Select** a certain timeframe | `ALT` + `mouse selection` | `⌘` + `mouse selection` (macOS) | | +| **Pan** forward or back in time | `click and drag` | `touch and drag` (touchpad/touchscreen) | | +| **Zoom** to a specific timeframe | `SHIFT` + `mouse selection` | | | +| **Zoom** in/out | `SHIFT`/`ALT` + `mouse scrollwheel` | `SHIFT`/`ALT` + `two-finger pinch` (touchpad/touchscreen) | `SHIFT`/`ALT` + `two-finger scroll` (touchpad/touchscreen) | + +Here's how chart synchronization looks while zooming and panning: + +![Animated GIF of the standard Netdata dashboard being manipulated and synchronizing +charts](https://user-images.githubusercontent.com/1153921/80839230-b034a800-8baf-11ea-9cb2-99c1e10f0f85.gif) + +You can also perform all these actions using the small +rewind/play/fast-forward/zoom-in/zoom-out buttons that appear in the +bottom-right corner of each chart. + +Additionally, resize charts by clicking-and-dragging the icon on the bottom-right corner of any chart. To restore the +chart to its original height, double-click the same icon. + +![Animated GIF of resizing a chart and resetting it to the default +height](https://user-images.githubusercontent.com/1153921/80842459-7d41e280-8bb6-11ea-9488-1bc29f94d7f2.gif) + +## Charts, contexts, families + +Before customizing the standard web dashboard, creating a custom dashboard, +configuring an alarm, or writing a collector, it's crucial to understand how +Netdata organizes metrics into charts, dimensions, families, and contexts. + +### Charts + +A **chart** is an individual, interactive, always-updating graphic displaying +one or more collected/calculated metrics. Charts are generated by +[collectors](/collectors/README.md). + +Here's the system CPU chart, the first chart displayed on the standard +dashboard: + +![Screenshot of the system CPU chart in the Netdata +dashboard](https://user-images.githubusercontent.com/1153921/62720972-0b8a8e80-b9c0-11e9-930b-4829f7b17cfd.png) + +Netdata displays a chart's name in parentheses above the chart. For example, if +you navigate to the system CPU chart, you'll see the label: **Total CPU +utilization (system.cpu)**. In this case, the chart's name is `system.cpu`. +Netdata derives the name from the chart's [context](#contexts). + +### Dimensions + +A **dimension** is a value that gets shown on a chart. The value can be raw data +or calculated values, such as percentages, aggregates, and more. + +Charts are capable of showing more than one dimension. Netdata shows these +dimensions on the right side of the chart, beneath the date and time. Again, the +`system.cpu` chart will serve as a good example. + +![Screenshot of the dimensions shown in the system CPU chart in the Netdata +dashboard](https://user-images.githubusercontent.com/1153921/62721031-2bba4d80-b9c0-11e9-9dca-32403617ce72.png) + +Here, the `system.cpu` chart is showing many dimensions, such as `user`, +`system`, `softirq`, `irq`, and more. + +Note that other applications sometimes use the word _series_ instead of +_dimension_. + +### Families + +A **family** is _one_ instance of a monitored hardware or software resource that +needs to be monitored and displayed separately from similar instances. + +For example, if your system has multiple disk drives at `sda` and `sdb`, Netdata +will put each interface into their own family. Same goes for software resources, +like multiple MySQL instances. We call these instances "families" because the +charts associated with a single disk instance, for example, are often related to +each other. Relatives, family... get it? + +When relevant, Netdata prefers to organize charts by family. When you visit the +**Disks** section, you will see your disk drives organized into families, and +each family will have one or more charts: `disk`, `disk_ops`, `disk_backlog`, +`disk_util`, `disk_await`, `disk_avgsz`, `disk_svctm`, `disk_mops`, and +`disk_iotime`. + +In the screenshot below, the disk family `sdb` shows a few gauges, followed by a +few of the associated charts: + +![Screenshot of a disk drive family and associated charts in the Netdata +dashboard](https://user-images.githubusercontent.com/1153921/62721362-e34f5f80-b9c0-11e9-8d2e-9a3bec48e920.png) + +Netdata also creates separate submenu entries for each family in the right +navigation page so you can easily navigate to the instance you're interested in. +Here, Netdata has made several submenus under the **Disk** menu. + +![Screenshot of the disks menu and +submenus](https://user-images.githubusercontent.com/1153921/62721531-3cb78e80-b9c1-11e9-89c2-fdd736aec7d4.png) + +### Contexts + +A **context** is a way of grouping charts by the types of metrics collected and +dimensions displayed. Different charts with the same context will show the same +dimensions, but for different instances (families) of hardware/software +resources. + +For example, the **Disks** section will often use many contexts (`disk.io`, +`disk.ops`, `disk.backlog`, `disk.util`, and so on). Netdata then creates an +individual chart for each context, and groups them by family. + +Netdata names charts according to their context according to the following +structure: `[context].[family]`. A chart with the `disk.util` context, in the +`sdb` family, gets the name `disk_util.sdb`. Netdata shows that name in the +top-left corner of a chart. + +Given the four example contexts, and two families of `sdb` and `sdd`, Netdata +will create the following charts and their names: + +Context | `sdb` family | `sdd` family +--- | --- | --- +`disk.io` | `disk_io.sdb` | `disk_io.sdd` +`disk.ops` | `disk_ops.sdb` | `disk_ops.sdd` +`disk.backlog` | `disk_backlog.sdb` | `disk_backlog.sdd` +`disk.util` | `disk_util.sdb` | `disk_util.sdd` + +And here's what two of those charts in the `disk.io` context look like under +`sdb` and `sdd` families: + +![context_01](https://user-images.githubusercontent.com/1153921/62728232-177e4c80-b9d0-11e9-9e29-2a6c59d4d873.png) +![context_02](https://user-images.githubusercontent.com/1153921/62728234-1b11d380-b9d0-11e9-8904-07befd8ac592.png) + +As you can see in the screenshot, you can view the context of a chart if you +hover over the date above the list of dimensions. A tooltip will appear that +shows you two pieces of information: the collector that produces the chart, and +the chart's context. + +Netdata also uses [contexts for alarm templates](/health/REFERENCE.md#alarm-line-on). You can create an alarm for the +`net.packets` context to receive alerts for any chart with that context, no matter which family it's attached to. + +## Positive and negative values on charts + +To improve clarity on charts, Netdata dashboards present **positive** values for +metrics representing `read`, `input`, `inbound`, `received` and **negative** +values for metrics representing `write`, `output`, `outbound`, `sent`. + +![Screenshot showing positive and negative +values](https://user-images.githubusercontent.com/1153921/81870401-9d649080-952a-11ea-80e3-4a7b480252ee.gif) + +_Netdata charts showing the bandwidth and packets of a network interface. +`received` is positive and `sent` is negative._ + +## Autoscaled y-axis + +Netdata charts automatically zoom vertically, to visualize the variation of each +metric within the visible timeframe. + +![Animated GIF showing the auso-scaling Y +axis](https://user-images.githubusercontent.com/1153921/80838276-8084a080-8bad-11ea-8167-8d5ab2fb1be1.gif) + +_A zero-based `stacked` chart, automatically switches to an auto-scaled `area` +chart when a single dimension is selected._ + +## dashboard.js + +Netdata uses the `dashboards.js` file to define, configure, create, and update +all the charts and other visualizations that appear on any Netdata dashboard. +You need to put `dashboard.js` on any HTML page that's going to render Netdata +charts. + +The [custom dashboards documentation](/web/gui/custom/README.md) contains examples of such +custom HTML pages. + +### Generating dashboard.js + +We build the `dashboards.js` file by concatenating all the source files located +in the `web/gui/src/dashboard.js/` directory. That's done using the provided +build script: + +```sh +cd web/gui +make +``` + +If you make any changes to the `src` directory when developing Netdata, you +should regenerate the `dashboard.js` file before you commit to the Netdata +repository. + +[![analytics](https://www.google-analytics.com/collect?v=1&aip=1&t=pageview&_s=1&ds=github&dr=https%3A%2F%2Fgithub.com%2Fnetdata%2Fnetdata&dl=https%3A%2F%2Fmy-netdata.io%2Fgithub%2Fweb%2FREADME&_u=MAC~&cid=5792dfd7-8dc4-476b-af31-da2fdb9f93d2&tid=UA-64295674-3)]() |