You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: content/en/tko/session-5/check-new-navigator-short.md
+35-26Lines changed: 35 additions & 26 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -8,34 +8,38 @@ weight: 2
8
8
9
9
The Kubernetes Navigator offers you two separate use cases to view your Kubernetes data.
10
10
11
-
* The `K8s workloads` is focusing on providing information in regards to workloads a.k.a. *your deployments*.
12
-
* The `K8s nodes` is focusing on providing insight into the performance of clusters, Nodes, Pods & Containers.
11
+
* The **K8s workloads** is focusing on providing information in regards to workloads a.k.a. *your deployments*.
12
+
* The **K8s nodes** is focusing on providing insight into the performance of clusters, Nodes, Pods & Containers.
13
13
14
-
You initially select either view depending on your need. (You can switch between the view on the fly if required) The most common one we will use in this workshop is the workload view and we will focus on that specifically.
14
+
You initially select either view depending on your need (you can switch between the view on the fly if required). The most common one we will use in this workshop is the workload view and we will focus on that specifically.
15
15
16
-
### 1.1 Finding your K8s cluster name
16
+
### 1.1 Finding your K8s Cluster name
17
17
18
-
Your first task is to identify and find your own cluster. The cluster will be named after your EC2 instance name: `ws-5-X-k3s-cluster`where `X` is the number of the EC2 instance assigned to you.
18
+
Your first task is to identify and find your own cluster. The cluster will be named after your EC2 instance name.
19
19
20
-
To find your node name, look at the prompt of your EC2 instance. For example, if you are assigned the 7th EC2 instance, the prompt will show:`ubuntu@ws-5-7 ~ $`
20
+
To confirm your EC2 instance name, look at the prompt of your EC2 instance. For example, if you are assigned the 7th EC2 instance, the prompt will show:
21
21
22
-
This means your cluster is named: `ws-5-7-k3s-cluster`. Please make a note of your cluster name as you will need this later in the workshop for filtering.
22
+
``` bash
23
+
ubuntu@emea-ws-7 ~ $
24
+
```
25
+
26
+
This means your cluster is named: `emea-ws-7-k3s-cluster`. Please make a note of your cluster name as you will need this later in the workshop for filtering.
23
27
24
28
## 2. Workloads & Workload Details Pane
25
29
26
-
Go to the **Infrastructure** menu item in the Observability UI and select **Kubernetes**. Go to the **Infrastructure** page in the Observability UI and select **Kubernetes**, this will offer you a set of Kubernetes services, one of them being the `K8s workloads` pane
30
+
Go to the **Infrastructure** menu item in the Observability UI and select **Kubernetes**. Go to the **Infrastructure** page in the Observability UI and select **Kubernetes**, this will offer you a set of Kubernetes services, one of them being the **K8s workloads** pane.
27
31
28
32
The pane will show a tiny graph giving you a bird's eye view of the load being handled across those Workloads. Also, if there are any alerts for one of the workloads, you will see a small alert indicator as shown in the image above.
29
33
30
34
31
35
32
-
Click on the `K8s workloads` pane and you will be taken to the workload view.
36
+
Click on the **K8s workloads** pane and you will be taken to the workload view.
33
37
34
-
Initially, you will see all the workloads that are reported by your clusters into your Observability Cloud Org. If an alert has fired for any of the workloads,it will be highlighted on the top right, *as marked with a red stripe in the image below*. You can go directly to the alert by clicking it to expand it.
38
+
Initially, you will see all the workloads that are reported by your clusters into your Observability Cloud Org. If an alert has fired for any of the workloads,it will be highlighted on the top right *as marked with a red stripe* in the image below. You can go directly to the alert by clicking it to expand it.
35
39
36
40
37
41
38
-
Now, let's find your own cluster by filtering on the field `k8s.cluster.name` in the filter toolbar, *as marked with a blue stripe*. Note: you can enter a partial name into the search box, such as 'ws-5-7*', to quickly find your Cluster. Also it's a very good idea to switch the default time from the default 3 hours back to 15 minutes.
42
+
Now, let's find your own cluster by filtering on the field `k8s.cluster.name` in the filter toolbar, *as marked with a blue stripe*. Note: you can enter a partial name into the search box, such as 'emea-ws-7*', to quickly find your Cluster. Also it's a very good idea to switch the default time from the default **-3h** back to last 15 minutes (**-15m**).
39
43
40
44
You should now just see information for your own cluster.
41
45
@@ -45,31 +49,36 @@ How many workloads are running & how many namespaces are in your Cluster?
45
49
46
50
### 2.1 Using the Navigator Selection chart
47
51
48
-
The `k8s workloads` Table is a common feature used across most of the Navigator's and will offer you a list view of the data you are viewing. In our case, it shows a list of `Pods Failed` grouped by `k8s.namespace.name`. You will find this pane used in various navigators in this workshop (ex: Workloads & Apache Servers). The way the selection pane works is similar across the navigators in that the information shown will match the selection criteria in your filters.
52
+
The **K8s workloads** table is a common feature used across most of the Navigator's and will offer you a list view of the data you are viewing. In our case, it shows a list of `Pods Failed` grouped by `k8s.namespace.name`.
53
+
54
+
You will find this pane used in various Navigators in this workshop (ex: Workloads & Apache Servers). The way the selection pane works is similar across the Navigators in that the information shown will match the selection criteria in your filters.
Now let's change the list view to a heat map view by selecting either the Heat map icon or List icon in the upper-right corner of the screen: [heat-map-toggle](../images/heatmaptoggle.png) *(as marked with a purple line)*
58
+
Now let's change the list view to a heat map view by selecting either the Heat map icon or List icon in the upper-right corner of the screen*(as marked with a purple line)*.
53
59
54
60
Changing this option will result in the following representation, which is one we will use most of the time in this workshop:
55
61
56
62
57
63
58
-
In this view, you will note that each workload is now a colored rectangle. These rectangles change color according to the `color by` option you selected, *as marked by a green line*. The colors give a visual indication of health and/or usage. (You can check the meaning by clicking on `legend !`)
64
+
In this view, you will note that each workload is now a colored rectangle. These rectangles change color according to the **Color by** option you selected, *as marked by a green line* above. The colors give a visual indication of health and/or usage. You can check the meaning by hovering over the **legend** exclamation icon {{% icon icon="exclamation-circle" %}}
65
+
bottom right of the heatmaps.
59
66
60
-
Another valuable option in this screen is `Find Outliers` which provides historical analytics of your clusters based on what is selected in the `Color by` box.
67
+
Another valuable option in this screen is **Find Outliers** which provides historical analytics of your clusters based on what is selected in the **Color by** dropdown.
61
68
62
-
Now, let's select the `File system usage (bytes)` from the **Color by** drop down box, *as marked with a green line* then click on the `Find outliers` drop down *as marked by a yellow stripe in the above image* and make sure you change the Strategy in the dialog to `Deviation from Median` as below:
69
+
Now, let's select the **File system usage (bytes)** from the **Color by** drop down box, *as marked with a green line* then click on the **Find outliers** drop down *as marked by a yellow stripe* in the above image and make sure you change the Strategy in the dialog to **Deviation from Median** as below:
The `Find outliers` view is very useful when you need to view a selection of your workloads (or any service depending on the navigator used) and quickly need to figure out if something has changed. It will give you fast insight into items (workloads in our case) that are performing differently (both increased or decreased) which helps to make it easier to spot problems.
77
+
The **Find outliers** view is very useful when you need to view a selection of your workloads (or any service depending on the Navigator used) and quickly need to figure out if something has changed.
71
78
72
-
Now switch the `Find Outliers` option back to off.
79
+
It will give you fast insight into items (workloads in our case) that are performing differently (both increased or decreased) which helps to make it easier to spot problems.
80
+
81
+
Now switch the **Find Outliers** option back to off.
73
82
74
83
### 2.2 The Workload Overview pane
75
84
@@ -87,14 +96,13 @@ You can expand the Workload name by hovering your mouse on it, in case the name
To filter to a specific Workload, you can click on three dots `...` next to the workload name in the *k8s.workload.name* column and choose `filter` from the drop down box.
99
+
To filter to a specific workload, you can click on three dots `...` next to the workload name in the **k8s.workload.name** column and choose **Filter** from the dropdown box.
This will add the selected workload to your filters. Try this for the *splunk-otel-collector-k8s-cluster-receiver* workload. It should give you a single rectangle in the `splunk' namespace.
95
-
Confirm that you have the right filter by hovering over the rectangle, and if it is indeed the *splunk-otel-collector-k8s-cluster-receiver* workload, double click on it.
103
+
This will add the selected workload to your filters. Try this for the **splunk-otel-collector-k8s-cluster-receiver** workload. It should give you a single rectangle in the `splunk` namespace.
96
104
97
-
This brings you to a more detailed view of your workload.
105
+
Confirm that you have the right filter by hovering over the rectangle, and if it is indeed the **splunk-otel-collector-k8s-cluster-receiver** workload, by double-clicking on it. This brings you to a more detailed view of your workload.
What are the CPU request & CPU limit units for the otel-collector?
@@ -104,11 +112,12 @@ At this point you can drill into the information of the pods, but that is outsid
104
112
105
113
## 3. Pivot Sidebar
106
114
107
-
Later in the workshop, you will deploy an Apache server into your cluster which will cause a `pivot bar` to appear.
108
-
As soon as any metric that is used by a Navigator, is flowing into your Observability org, the system will add that service to the pivot bar.
115
+
Later in the workshop, you will deploy an Apache server into your cluster which will create a new icon in the **Pivot Sidebar**.
116
+
117
+
As soon as any metric that is used by a Navigator, is flowing into your Observability org, the system will add that service to the Pivot Sidebar.
109
118
110
-
The pivot bar will expand and a link to the discovered service will be added as seen in the image below:
119
+
The Pivot Sidebar will expand and a link to the discovered service will be added as seen in the image below:
111
120
112
121
113
122
114
-
This will allow for easy switching between Navigators. The same applies for your Apache server instance, it will have a pivot bar allowing you to quickly jump back to the Kubernetes navigator.
123
+
This will allow for easy switching between Navigators. The same applies for your Apache server instance, it will have a Pivot Sidebar allowing you to quickly jump back to the Kubernetes navigator.
As the new Kubernetes Navigator is still in Preview, some steps of this workshop might not work as expected. If you encounter any issues, please do the following:
11
+
12
+
- Switch back to the old Kubernetes Navigator by clicking on the big blue {{% button style="blue" %}}Switch to old navigator{{% /button %}} button in the top right corner of the Kubernetes Navigator.
13
+
14
+
- Let us know in the **#tko-2023-o11y-session-5 channel** in Slack.
15
+
{{% /notice %}}}
16
+
9
17
As we are in the process of switching to the new generation of the Kubernetes Navigator, please check if you are already on the new Kubernetes navigator.
10
18
11
19
When you select **Infrastructure** from the main menu on the left, followed by selecting **Kubernetes**, you should see two services panes for Kubernetes, similar like the ones below:
@@ -45,7 +53,7 @@ Most of our customers will make use of some kind of private or public cloud serv
45
53
46
54
Namespaces are a way to organize these large Kubernetes clusters into virtual sub-clusters. This can be helpful when different teams or projects share a Kubernetes cluster as this will give them the easy ability to just see and work with their own stuff.
47
55
48
-
Any number of namespaces are supported within a cluster, each logically separated from others but with the ability to communicate with each other. Components are only "visible" when selecting a namespace or when adding the `--all-namespaces` flag to `kubectl` instead of allowing you to view just the components relevant to your project by selecting your namespace.
56
+
Any number of namespaces are supported within a cluster, each logically separated from others but with the ability to communicate with each other. Components are only **visible** when selecting a namespace or when adding the `--all-namespaces` flag to `kubectl` instead of allowing you to view just the components relevant to your project by selecting your namespace.
49
57
50
58
Most customers will want to install the Splunk OpenTelemetry Collector in a separate namespace. This workshop will follow that practice.
0 commit comments