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
This article provides troubleshooting information to help you with problems you might have experienced when you tried to enable or use VM insights.
13
+
14
+
This article provides troubleshooting information to help you with problems that you might experience when you try to enable or use virtual machine (VM) insights in Azure Monitor.
14
15
15
16
## Can't enable VM insights on a machine
16
-
When you onboard an Azure virtual machine from the Azure portal, the following steps occur:
17
17
18
-
- A default Log Analytics workspace is created if that option was selected.
19
-
- Azure Monitor Agent is installed on Azure virtual machines by using a VM extension if the agent is already installed.
20
-
- The Dependency agent is installed on Azure virtual machines by using an extension if it's required.
18
+
When you onboard an Azure virtual machine from the Azure portal, the following actions occur:
19
+
20
+
- A default Log Analytics workspace is created if you selected that option.
21
+
- The Azure Monitor Agent is installed on Azure virtual machines through a VM extension, if the agent is already installed.
22
+
- The Dependency Agent is installed on Azure virtual machines through an extension, if it's required.
21
23
22
24
During the onboarding process, each of these steps is verified and a notification status appears in the portal. Configuration of the workspace and the agent installation typically takes 5 to 10 minutes. It takes another 5 to 10 minutes for data to become available to view in the portal.
23
25
24
-
If you receive a message that the virtual machine needs to be onboarded after you've performed the onboarding process, allow for up to 30 minutes for the process to finish. If the issue persists, see the following sections for possible causes.
26
+
If you receive a message that the virtual machine needs to be onboarded after you perform the onboarding process, allow up to 30 minutes for the process to finish. If the issue persists, see the following sections for possible causes.
25
27
26
28
### Is the virtual machine running?
27
-
If the virtual machine has been turned off for a while, is off currently, or was only recently turned on, you won't have data to display until data arrives.
29
+
30
+
If the virtual machine has been turned off for a while, is off currently, or was only recently turned on, you won't have data to display until data arrives.
28
31
29
32
### Is the operating system supported?
30
-
If the operating system isn't in the list of [supported operating systems](vminsights-enable-overview.md#supported-operating-systems), the extension fails to install and you see a message that we're waiting for data to arrive.
33
+
34
+
If the operating system isn't in the [list of supported operating systems](vminsights-enable-overview.md#supported-operating-systems), installation of the extension fails and you get a message about waiting for data to arrive.
31
35
32
36
> [!IMPORTANT]
33
-
> Post April 11, 2022, if you aren't seeing your virtual machine in the VM insights solution, you might be running an older version of the Dependency agent. For more information, see the blog post [Potential breaking changes for VM insights Linux customers](https://techcommunity.microsoft.com/t5/azure-monitor-status/potential-breaking-changes-for-vm-insights-linux-customers/ba-p/3271989). Not applicable for Windows machines and before April 11, 2022.
37
+
> As of April 11, 2022, if your virtual machine doesn't appear in VM insights, you might be running an older version of the Dependency Agent. For more information, see the blog post [Potential breaking changes for VM insights Linux customers](https://techcommunity.microsoft.com/t5/azure-monitor-status/potential-breaking-changes-for-vm-insights-linux-customers/ba-p/3271989). This consideration doesn't apply for Windows machines and for VMs that you onboarded before April 11, 2022.
38
+
39
+
### Was the extension installed properly?
34
40
35
-
### Did the extension install properly?
36
-
If you still see a message that the virtual machine needs to be onboarded, it might mean that one or both of the extensions failed to install correctly. Check the **Extensions** page for your virtual machine in the Azure portal to verify that the following extensions are listed.
41
+
If you still see a message that the virtual machine needs to be onboarded, it might mean that one or both of the extensions failed to be installed correctly. In the Azure portal, on the **Extensions** pane for your virtual machine, check whether the following extensions appear.
37
42
38
-
| Operating system | Agents |
43
+
| Operating system | Agents |
39
44
|:---|:---|
40
45
| Windows | MicrosoftMonitoringAgent<br>Microsoft.Azure.Monitoring.DependencyAgent |
41
46
| Linux | OMSAgentForLinux<br>DependencyAgentLinux |
42
47
43
-
If you don't see both the extensions for your operating system in the list of installed extensions, they must be installed. If the extensions are listed but their status doesn't appear as *Provisioning succeeded*, remove the extensions and reinstall them.
48
+
If you don't see both extensions for your operating system in the list of installed extensions, you must install them. If the extensions are listed but their status doesn't appear as **Provisioning succeeded**, remove the extensions and reinstall them.
44
49
45
50
### Do you have connectivity issues?
46
-
For Windows machines, you can use the TestCloudConnectivity tool to identify connectivity issue. This tool is installed by default with the agent in the folder *%SystemDrive%\Program Files\Microsoft Monitoring Agent\Agent*. Run the tool from an elevated command prompt. It returns results and highlights where the test fails.
47
-
<!-- convertborder later -->
48
-
:::image type="content" source="media/vminsights-troubleshoot/test-cloud-connectivity.png" lightbox="media/vminsights-troubleshoot/test-cloud-connectivity.png" alt-text="Screenshot that shows the TestCloudConnectivity tool." border="false":::
49
51
50
-
### More agent troubleshooting
52
+
For Windows machines, you can use the TestCloudConnectivity tool to identify the connectivity issue. This tool is installed by default with the agent in the folder *%SystemDrive%\Program Files\Microsoft Monitoring Agent\Agent*. Run the tool from an elevated command prompt. It returns results and highlights where the test fails.
51
53
52
-
See the following articles for troubleshooting issues with the Log Analytics agent:
54
+
:::image type="content" source="media/vminsights-troubleshoot/test-cloud-connectivity.png" lightbox="media/vminsights-troubleshoot/test-cloud-connectivity.png" alt-text="Screenshot that shows the tool for testing cloud connectivity." border="false":::
53
55
54
-
-[Troubleshoot issues with the Log Analytics agent for Windows](../agents/agent-windows-troubleshoot.md)
55
-
56
-
-[Troubleshoot issues with the Log Analytics agent for Linux](../agents/agent-linux-troubleshoot.md)
56
+
## DCR created by the VM insights process was modified and now data is missing
57
57
58
-
##DCR created by VMInsight Process was modified and now data is missing
58
+
### Identify the issue
59
59
60
-
### Identifying the Issue
60
+
To identify whether a data collection rule (DCR) was modified:
61
61
62
-
To identify if this is the case you would first browse to the Monitor Dashboard, locate the Data Collection Rule in Question and view the JSON properties using the link on the top right-hand side of the overview blade:
62
+
1. Go to the Azure Monitor dashboard.
63
+
1. Locate the DCR.
64
+
1. View the JSON properties by using the **JSON View** link on the upper-right side of the overview pane.
63
65
64
-
66
+
65
67
66
-
You will see that stream name has been changed from its original name to reflect the performance counter stream name:
68
+
In this example, the original stream name was changed to reflect the name of the performance counter stream.
67
69
68
-
70
+
69
71
70
-
We can see while the counter sections are pointing the perf table, the stream dataflow is still configured for the proper destination Microsof-InsightsMetrics.
72
+
Although the counter sections are pointing the `perf` table, the stream dataflow is still configured for the proper destination `Microsoft-InsightsMetrics`.
71
73
72
-
### Resolving the Issue
74
+
### Resolve the issue
73
75
74
-
This issue can't be resolved using the Monitor Dashboard directly, but we can fix this by exporting the template, normalizing the name and then import the modified rule over itself.
76
+
You can't resolve this issue by using the Azure Monitor dashboard directly. But you can resolve it by exporting the template, normalizing the name, and then importing the modified rule over itself.
75
77
76
78
#### Export the DCR and save locally
77
79
78
-
1. To do this first we must export the DCR:
79
-
80
-
80
+
1. On the Azure Monitor dashboard, go to the DCR.
81
+
82
+
2. Select **Export template**.
83
+
84
+
81
85
82
-
2. After selecting the Export Template blade for the selected DCR the portal will create the template file and a matching parameter file. Once this is complete, we can download the template package save and save it locally.
86
+
3. On the **Export template** pane for the selected DCR, the portal creates the template file and a matching parameter file. After this process is complete, select **Download** to download the template package and save it locally.
83
87
84
-
88
+
85
89
86
-
3.Open the file
90
+
4. Select **Open file**.
87
91
88
-
92
+
89
93
90
-
4. Copy these to a local folder:
94
+
5. Copy the two .json files to a local folder.
91
95
92
-
96
+
93
97
94
-
#### Modify the Template
98
+
#### Modify the template
95
99
96
-
1. Open the template file in the editor of your choice and locate the invalid stream name under the performance counter data source.
97
-
98
-
100
+
1. Open the template file in the editor of your choice. Then locate the invalid stream name under the performance counter data source.
99
101
100
-
2. Using the valid stream name from the dataflow node fix the invalid reference, then save and close your file:
101
-
102
-
102
+
103
103
104
-
#### Import the Template using the Custom Deployment Feature
104
+
2. Using the valid stream name from the dataflow node, fix the invalid reference. Then save and close your file.
105
105
106
-
1. Back in the portal, search for and navigate to the custom template deployment:
106
+
107
107
108
-
108
+
#### Import the template by using the custom deployment feature
109
109
110
-
2. Choose the Option to "Build Your Own Template"
110
+
1. Back in the portal, search for and then select **Deploy a custom template**.
111
111
112
-
112
+
113
113
114
-
3. Using the "Load File" link browse to your saved template and parameter file:
114
+
2. On the **Custom deployment** pane, select **Build your own template in the editor**.
115
115
116
-
116
+
117
117
118
-
4. Visually inspect the template to validate the change is in place and select the Save button
118
+
3. Select **Load file**, and then browse to your saved template and parameter file.
119
119
120
-
120
+
121
+
122
+
4. Visually inspect the template to validate the change is in place and select the Save button.
123
+
124
+
121
125
122
126
5. From here the portal will use the parameter file to fill in the deployment options (which can be changed) or left intact to overwrite the existing DCR, Once completed select the review and Create button.
123
127
124
-
128
+
125
129
126
130
6. After validation then we can select the Create button to finalize the deployment.
127
131
128
-
132
+
129
133
130
134
7. After the deployment is complete, we can browse to the DCR again and review the JSON in overview blade:
131
135
132
-
136
+
133
137
134
138
8. The agent will detect this change and download the new configuration, and this should restore ingestion to the insight metrics table.
135
139
@@ -138,6 +142,7 @@ This issue can't be resolved using the Monitor Dashboard directly, but we can fi
138
142
If the agents appear to be installed correctly but you don't see any data in the **Performance** view, see the following sections for possible causes.
139
143
140
144
### Has your Log Analytics workspace reached its data limit?
145
+
141
146
Check the [capacity reservations and the pricing for data ingestion](https://azure.microsoft.com/pricing/details/monitor/).
142
147
143
148
### Is your virtual machine agent connected to Azure Monitor Logs?
@@ -156,25 +161,31 @@ If you don't see any data or if the computer hasn't sent a heartbeat recently, y
156
161
157
162
See the following sections for issues with the **Map** view.
158
163
159
-
### Is the Dependency agent installed?
160
-
Use the information in the preceding sections to determine if the Dependency agent is installed and working properly.
164
+
### Is the Dependency Agent installed?
165
+
166
+
Use the information in the preceding sections to determine if the Dependency Agent is installed and working properly.
161
167
162
168
### Are you on the Log Analytics free tier?
169
+
163
170
The [Log Analytics free tier](https://azure.microsoft.com/pricing/details/monitor/) is a legacy pricing plan that allows for up to five unique Service Map machines. Any subsequent machines won't appear in Service Map, even if the prior five are no longer sending data.
164
171
165
172
### Is your virtual machine sending log and performance data to Azure Monitor Logs?
173
+
166
174
Use the log query in the [Performance view has no data](#performance-view-has-no-data) section to determine if data is being collected for the virtual machine. If no data is being collected, use the TestCloudConnectivity tool to determine if you have connectivity issues.
167
175
168
176
## Virtual machine appears in the Map view but has missing data
169
-
If the virtual machine is in the **Map** view, the Dependency agent is installed and running, but the kernel driver didn't load, check the log file at the following locations:
170
177
171
-
| Operating system | Log |
178
+
If the virtual machine is in the **Map** view, the Dependency Agent is installed and running, but the kernel driver didn't load, check the log file at the following locations:
179
+
180
+
| Operating system | Log |
172
181
|:---|:---|
173
182
| Windows | C:\Program Files\Microsoft Dependency Agent\logs\wrapper.log |
174
183
| Linux | /var/opt/microsoft/dependency-agent/log/service.log |
175
184
176
185
The last lines of the file should indicate why the kernel didn't load. For example, the kernel might not be supported on Linux if you updated your kernel.
177
186
178
-
## Next steps
187
+
## Related content
179
188
180
-
For more information on onboarding VM insights agents, see [Enable VM insights overview](vminsights-enable-overview.md).
189
+
-[Enable VM insights overview](vminsights-enable-overview.md).
190
+
-[Troubleshoot issues with the Log Analytics agent for Windows](../agents/agent-windows-troubleshoot.md)
191
+
-[Troubleshoot issues with the Log Analytics agent for Linux](../agents/agent-linux-troubleshoot.md)
0 commit comments