Update overview.md (#14607)
Chris Akritidis committed
Feb 25, 2023 at 08:07 UTC
7c124972e17f73692e464c529de6be27f7236166
1 file changed
+23
-219
docs/cloud/visualize/overview.md
+23
-219
@@ -1,250 +1,54 @@
1
-<!--
2
-title: "Home, Overview and Single Node view"
3
-description: >-
4
- "The Home tab automatically presents relevant information of your War Room, the Overview uses composite
5
- charts from all the nodes in a given War Room and Single Node view provides a look at a specific Node"
6
-custom_edit_url: "https://github.com/netdata/netdata/blob/master/docs/cloud/visualize/overview.md"
7
-sidebar_label: "Home, Overview and Single Node view"
8
-learn_status: "Published"
9
-learn_topic_type: "Concepts"
10
-learn_rel_path: "Operations/Visualizations"
11
--->
1
+# Home, Overview and Single Node view
2
3
## Home
4
5
The Home tab provides a predefined dashboard of relevant information about entities in the War Room.
6
17
-This tab will
18
-automatically present summarized information in an easily digestible display. You can see information about your
7
+This tab will automatically present summarized information in an easily digestible display. You can see information about your
8
nodes, data collection and retention stats, alerts, users and dashboards.
9
21
-## Overview
10
+## Overview and Single Node view
11
12
The Overview tab is another great way to monitor infrastructure using Netdata Cloud. While the interface might look
24
-similar to local
25
-dashboards served by an Agent Overview uses **composite charts**.
13
+similar to local dashboards served by an Agent Overview uses **composite charts**.
14
These charts display real-time aggregated metrics from all the nodes (or a filtered selection) in a given War Room.
15
28
-With Overview's composite charts, you can see your infrastructure from a single pane of glass, discover trends or
29
-anomalies, then drill down by grouping metrics by node and jumping to single-node dashboards for root cause analysis.
16
+When you [interact with composite charts](https://github.com/netdata/netdata/blob/master/docs/cloud/visualize/interact-new-charts.md)
17
+you can see your infrastructure from a single pane of glass, discover trends or anomalies, and perform root cause analysis.
18
31
-## Single Node view
19
+The Single Node view dashboard is exactly the same as the Overview, but with a hard-coded filter to only show
20
+a single node.
21
33
-The Single Node view dashboard engine is the same as the Overview, meaning that it also uses **composite charts**, and
34
-displays real-time aggregated metrics from a specific node.
22
+### Chart Navigation Menu
23
36
-As mentioned above, the interface is similar to local dashboards served by an Agent but this dashboard also uses *
37
-*composite charts** which, in the case of a single node, will aggregate
38
-multiple chart _instances_ belonging to a context into a single chart. For example, on `disk.io` context it will get
39
-into a single chart an aggregated view of each disk the node has.
40
-
41
-Further tools provided in composite chart [definiton bar](/docs/cloud/visualize/overview#definition-bar) will allow you
42
-to explore in more detail what is happening on each _instance_.
43
-
44
-## Before you get started
45
-
46
-Only nodes with v1.25.0-127 or later of the the [open-source Netdata](https://github.com/netdata/netdata) monitoring
47
-agent can contribute to composite charts. If your node(s) use an earlier version of Netdata, you will see them marked as
48
-**needs upgrade** in various dropdowns.
49
-
50
-See our [update docs](https://github.com/netdata/netdata/blob/master/packaging/installer/UPDATE.md) for the preferred
51
-update method based on how you installed
52
-Netdata.
53
-
54
-## Composite charts
55
-
56
-The Overview uses composite charts, which aggregate metrics from all the nodes (or a filtered selection) in a given War
57
-Room.
58
-
59
-## Definition bar
60
-
61
-Each composite chart has a definition bar to provide information about the following:
62
-
63
-* Grouping option
64
-* Aggregate function to be applied in case multiple data sources exist
65
-* Instances
66
-* Nodes
67
-* Dimensions, and
68
-* Aggregate function over time to be applied if one point in the chart consists of multiple data points aggregated
69
-
70
-### Group by dimension, node, or chart
71
-
72
-Click on the **dimension** dropdown to change how a composite chart groups metrics.
73
-
74
-The default option is by _dimension_, so that each line/area in the visualization is the aggregation of a single
75
-dimension.
76
-This provides a per dimension view of the data from all the nodes in the War Room, taking into account filtering
77
-criteria if defined.
78
-
79
-A composite chart grouped by _node_ visualizes a single metric across contributing nodes. If the composite chart has
80
-five
81
-contributing nodes, there will be five lines/areas. This is typically an absolute value of the sum of the dimensions
82
-over each node but there
83
-are some opinionated-but-valuable exceptions where a specific dimension is selected.
84
-Grouping by nodes allows you to quickly understand which nodes in your infrastructure are experiencing anomalous
85
-behavior.
86
-
87
-A composite chart grouped by _instance_ visualizes each instance of one software or hardware on a node and displays
88
-these as a separate dimension. By grouping the
89
-`disk.io` chart by _instance_, you can visualize the activity of each disk on each node that contributes to the
90
-composite
91
-chart.
92
-
93
-Another very pertinent example is composite charts over contexts related to cgroups (VMs and containers). You have the
94
-means to change the default group by or apply filtering to
95
-get a better view into what data your are trying to analyze. For example, if you change the group by to _instance_ you
96
-get a view with the data of all the instances (cgroups) that
97
-contribute to that chart. Then you can use further filtering tools to focus the data that is important to you and even
98
-save the result to your own dashboards.
99
-
100
-
101
-
102
-### Aggregate functions over data sources
103
-
104
-Each chart uses an opinionated-but-valuable default aggregate function over the data sources. For example,
105
-the `system.cpu` chart shows the
106
-average for each dimension from every contributing chart, while the `net.net` chart shows the sum for each dimension
107
-from every contributing chart, which can also come from multiple networking interfaces.
108
-
109
-The following aggregate functions are available for each selected dimension:
110
-
111
-- **Average**: Displays the average value from contributing nodes. If a composite chart has 5 nodes with the following
112
- values for the `out` dimension—`-2.1`, `-5.5`, `-10.2`, `-15`, `-0.1`—the composite chart displays a
113
- value of `−6.58`.
114
-- **Sum**: Displays the sum of contributed values. Using the same nodes, dimension, and values as above, the composite
115
- chart displays a metric value of `-32.9`.
116
-- **Min**: Displays a minimum value. For dimensions with positive values, the min is the value closest to zero. For
117
- charts with negative values, the min is the value with the largest magnitude.
118
-- **Max**: Displays a maximum value. For dimensions with positive values, the max is the value with the largest
119
- magnitude. For charts with negative values, the max is the value closet to zero.
120
-
121
-### Dimensions
122
-
123
-Select which dimensions to display on the composite chart. You can choose **All dimensions**, a single dimension, or any
124
-number of dimensions available on that context.
125
-
126
-### Instances
127
-
128
-Click on **X Instances** to display a dropdown of instances and nodes contributing to that composite chart. Each line in
129
-the
130
-dropdown displays an instance name and the associated node's hostname.
131
-
132
-### Nodes
133
-
134
-Click on **X Nodes** to display a dropdown of nodes contributing to that composite chart. Each line displays a hostname
135
-to help you identify which nodes contribute to a chart. You can also use this component to filter nodes directly on the
136
-chart.
137
-
138
-If one or more nodes can't contribute to a given chart, the definition bar shows a warning symbol plus the number of
139
-affected nodes, then lists them in the dropdown along with the associated error. Nodes might return errors because of
140
-networking issues, a stopped `netdata` service, or because that node does not have any metrics for that context.
141
-
142
-### Aggregate functions over time
143
-
144
-When the granularity of the data collected is higher than the plotted points on the chart an aggregation function over
145
-time
146
-is applied. By default the aggregation applied is _average_ but the user can choose different options from the
147
-following:
148
-
149
-* Min
150
-* Max
151
-* Average
152
-* Sum
153
-* Incremental sum (Delta)
154
-* Standard deviation
155
-* Median
156
-* Single exponential smoothing
157
-* Double exponential smoothing
158
-* Coefficient variation
159
-* Trimmed Median `*`
160
-* Trimmed Mean `*`
161
-* Percentile `**`
162
-
163
-:::info
164
-
165
-- `*` For **Trimmed Median and Mean** you can choose the percentage of data tha you want to focus on: 1%, 2%, 3%, 5%,
166
- 10%, 15%, 20% and 25%.
167
-- `**` For **Percentile** you can specify the percentile you want to focus on: 25th, 50th, 75th, 80th, 90th, 95th, 97th,
168
- 98th and 99th.
169
-
170
-:::
171
-
172
-For more details on each, you can refer to our Agent's HTTP API details
173
-on [Data Queries - Data Grouping](/docs/agent/web/api/queries#data-grouping).
174
-
175
-### Reset to defaults
176
-
177
-Click on the 3-dot icon (**⋮**) on any chart, then **Reset to Defaults**, to reset the definition bar to its initial
178
-state.
179
-
180
-## Jump to single-node dashboards
181
-
182
-Click on **X Charts**/**X Nodes** to display one of the two dropdowns that list the charts and nodes contributing to a
183
-given composite chart. For example, the nodes dropdown.
184
-
185
-
187
-
188
-To jump to a single-node dashboard, click on the link icon <img class="img__inline img__inline--link"
189
-src="https://user-images.githubusercontent.com/1153921/95762109-1d219300-0c62-11eb-8daa-9ba509a8e71c.png" /> next to the
190
-node you're interested in.
191
-
192
-The single-node dashboard opens in a new tab. From there, you can continue to troubleshoot or run [Metric
193
-Correlations](https://github.com/netdata/netdata/blob/master/docs/cloud/insights/metric-correlations.md) for faster root
194
-cause analysis.
195
-
196
-## Add composite charts to a dashboard
197
-
198
-Click on the 3-dot icon (**⋮**) on any chart, then click on **Add to Dashboard**. Click the **+** button for any
199
-dashboard you'd like to add this composite chart to, or create a new dashboard an initiate it with your chosen chart by
200
-entering the name and clicking **New Dashboard**.
201
-
202
-## Interacting with composite charts: pan, zoom, and resize
203
-
204
-You can interact with composite charts as you would with other Netdata charts. You can use the controls beneath each
205
-chart to pan, zoom, or resize the chart, or use various combinations of the keyboard and mouse. See
206
-the [chart interaction doc](https://github.com/netdata/netdata/blob/master/docs/dashboard/interact-charts.md) for
207
-details.
208
-
209
-## Menu
210
-
211
-The Overview uses a similar menu to local Agent dashboards and single-node dashboards in Netdata Cloud, with sections
24
+Netdata Cloud uses a similar menu to local Agent dashboards, with sections
25
and sub-menus aggregated from every contributing node. For example, even if only two nodes actively collect from and
26
monitor an Apache web server, the **Apache** section still appears and displays composite charts from those two nodes.
27
215
-
28
+
29
218
-One difference between the Overview's menu and those found in single-node dashboards or local Agent dashboards is that
30
+One difference between the Netdata Cloud menu and those found in local Agent dashboards is that
31
the Overview condenses multiple services, families, or instances into single sections, sub-menus, and associated charts.
32
221
-For services, let's say you have two concurrent jobs with the [web_log
222
-collector](https://github.com/netdata/go.d.plugin/blob/master/modules/weblog/README.md), one for Apache and another for
223
-Nginx. A single-node or
33
+For services, let's say you have two concurrent jobs with the
34
+[web_log collector](https://github.com/netdata/go.d.plugin/blob/master/modules/weblog/README.md), one for Apache and
35
+another for Nginx. A single-node or
36
local dashboard shows two section, **web_log apache** and **web_log nginx**, whereas the Overview condenses these into a
37
single **web_log** section containing composite charts from both jobs.
38
227
-The Overview also consdenses multiple families or multiple instances into a single **all** sub-menu and associated
39
+The Cloud also condenses multiple families or multiple instances into a single **all** sub-menu and associated
40
charts. For example, if Node A has 5 disks, and Node B has 3, each disk contributes to a single `disk.io` composite
229
-chart. The utility bar should show that there are 8 charts from 2 nodes contributing to that chart.
230
-
231
-This action applies to disks, network devices, and other metric types that involve multiple instances of a piece of
232
-hardware or software. The Overview currently does not display metrics from filesystems. Read more about [families and
233
-instances](https://github.com/netdata/netdata/blob/master/docs/dashboard/dimensions-contexts-families.md)
41
+chart. The utility bar should show that there are 8 charts from 2 nodes contributing to that chart. The aggregation applies
42
+to disks, network devices, and other metric types that involve multiple instances of a piece of hardware or software.
43
44
## Persistence of composite chart settings
45
46
+Of course you can
47
+[change the filtering or grouping](https://github.com/netdata/netdata/blob/master/docs/cloud/visualize/interact-new-charts.md)
48
+of metrics in the composite charts that aggregate all these instances, to see only the information you are interested in,
49
+and save that view in a custom dashboard.
50
+
51
When you change a composite chart via its definition bar, Netdata Cloud persists these settings in a query string
52
attached to the URL in your browser. You can "save" these settings by bookmarking this particular URL, or share it with
53
colleagues by having them copy-paste it into their browser.
54
241
-## What's next?
242
-
243
-For another way to view an infrastructure from a high level, see
244
-the [Nodes view](https://github.com/netdata/netdata/blob/master/docs/cloud/visualize/nodes.md).
245
-
246
-If you need a refresher on how Netdata's charts work, see our doc
247
-on [interacting with charts](https://github.com/netdata/netdata/blob/master/docs/dashboard/interact-charts.md).
248
-
249
-Or, get more granular with configuring how you monitor your infrastructure
250
-by [building new dashboards](https://github.com/netdata/netdata/blob/master/docs/cloud/visualize/dashboards.md).