Manage an Azure Stack Edge Pro GPU device via Windows PowerShell
APPLIES TO: 
Azure Stack Edge Pro - GPUAzure Stack Edge Pro 2Azure Stack Edge Pro RAzure Stack Edge Mini R
Azure Stack Edge Pro GPU solution lets you process data and send it over the network to Azure. This article describes some of the configuration and management tasks for your Azure Stack Edge Pro GPU device. You can use the Azure portal, local web UI, or the Windows PowerShell interface to manage your device.
This article focuses on how you can connect to the PowerShell interface of the device and the tasks you can do using this interface.
Connect to the PowerShell interface
Depending on the operating system of client, the procedures to remotely connect to the device are different.
Remotely connect from a Windows client
Prerequisites
Before you begin, make sure that:
Your Windows client is running Windows PowerShell 5.0 or later.
Your Windows client has the signing chain (root certificate) corresponding to the node certificate installed on the device. For detailed instructions, see Install certificate on your Windows client.
The
hostsfile located atC:\Windows\System32\drivers\etcfor your Windows client has an entry corresponding to the node certificate in the following format:<Device IP> <Node serial number>.<DNS domain of the device>Here is an example entry for the
hostsfile:10.100.10.10 1HXQG13.wdshcsso.com
Detailed steps
Follow these steps to remotely connect from a Windows client.
Run a Windows PowerShell session as an administrator.
Make sure that the Windows Remote Management service is running on your client. At the command prompt, type:
winrm quickconfigFor more information, see Installation and configuration for Windows Remote Management.
Assign a variable to the connection string used in the
hostsfile.$Name = "<Node serial number>.<DNS domain of the device>"Replace
<Node serial number>and<DNS domain of the device>with the node serial number and DNS domain of your device. You can get the values for node serial number from the Certificates page and DNS domain from the Device page in the local web UI of your device.To add this connection string for your device to the client’s trusted hosts list, type the following command:
Set-Item WSMan:\localhost\Client\TrustedHosts $Name -Concatenate -ForceStart a Windows PowerShell session on the device:
Enter-PSSession -ComputerName $Name -Credential ~\EdgeUser -ConfigurationName Minishell -UseSSLIf you see an error related to trust relationship, then check if the signing chain of the node certificate uploaded to your device is also installed on the client accessing your device.
Provide the password when prompted. Use the same password that is used to sign into the local web UI. The default local web UI password is Password1. When you successfully connect to the device using remote PowerShell, you see the following sample output:
Windows PowerShell Copyright (C) Microsoft Corporation. All rights reserved. PS C:\WINDOWS\system32> winrm quickconfig WinRM service is already running on this machine. PS C:\WINDOWS\system32> $Name = "1HXQG13.wdshcsso.com" PS C:\WINDOWS\system32> Set-Item WSMan:\localhost\Client\TrustedHosts $Name -Concatenate -Force PS C:\WINDOWS\system32> Enter-PSSession -ComputerName $Name -Credential ~\EdgeUser -ConfigurationName Minishell -UseSSL WARNING: The Windows PowerShell interface of your device is intended to be used only for the initial network configuration. Please engage Microsoft Support if you need to access this interface to troubleshoot any potential issues you may be experiencing. Changes made through this interface without involving Microsoft Support could result in an unsupported configuration. [1HXQG13.wdshcsso.com]: PS>
When you use the -UseSSL option, you are remoting via PowerShell over https. We recommend that you always use https to remotely connect via PowerShell. Within trusted networks, remoting via PowerShell over http is acceptable. You first enable remote PowerShell over http in the local UI. Then you can connect to PowerShell interface of the device by using the preceding procedure without the -UseSSL option.
If you are not using the certificates (we recommend that you use the certificates!), you can skip the certificate validation check by using the session options: -SkipCACheck -SkipCNCheck -SkipRevocationCheck.
$sessOptions = New-PSSessionOption -SkipCACheck -SkipCNCheck -SkipRevocationCheck
Enter-PSSession -ComputerName $Name -Credential ~\EdgeUser -ConfigurationName Minishell -UseSSL -SessionOption $sessOptions
Here is an example output when skipping the certificate check:
PS C:\WINDOWS\system32> $Name = "1HXQG13.wdshcsso.com"
PS C:\WINDOWS\system32> $sessOptions = New-PSSessionOption -SkipCACheck -SkipCNCheck -SkipRevocationCheck
PS C:\WINDOWS\system32> $sessOptions
MaximumConnectionRedirectionCount : 5
NoCompression : False
NoMachineProfile : False
ProxyAccessType : None
ProxyAuthentication : Negotiate
ProxyCredential :
SkipCACheck : True
SkipCNCheck : True
SkipRevocationCheck : True
OperationTimeout : 00:03:00
NoEncryption : False
UseUTF16 : False
IncludePortInSPN : False
OutputBufferingMode : None
MaxConnectionRetryCount : 0
Culture :
UICulture :
MaximumReceivedDataSizePerCommand :
MaximumReceivedObjectSize :
ApplicationArguments :
OpenTimeout : 00:03:00
CancelTimeout : 00:01:00
IdleTimeout : -00:00:00.0010000
PS C:\WINDOWS\system32> Enter-PSSession -ComputerName $Name -Credential ~\EdgeUser -ConfigurationName Minishell -UseSSL -SessionOption $sessOptions
WARNING: The Windows PowerShell interface of your device is intended to be used only for the initial network configuration. Please
engage Microsoft Support if you need to access this interface to troubleshoot any potential issues you may be experiencing.
Changes made through this interface without involving Microsoft Support could result in an unsupported configuration.
[1HXQG13.wdshcsso.com]: PS>
Important
In the current release, you can connect to the PowerShell interface of the device only via a Windows client. The -UseSSL option does not work with the Linux clients.
Create a support package
If you experience any device issues, you can create a support package from the system logs. Microsoft Support uses this package to troubleshoot the issues. Follow these steps to create a support package:
Use the
Get-HcsNodeSupportPackagecommand to create a support package. The usage of the cmdlet is as follows:Get-HcsNodeSupportPackage [-Path] <string> [-Zip] [-ZipFileName <string>] [-Include {None | RegistryKeys | EtwLogs | PeriodicEtwLogs | LogFiles | DumpLog | Platform | FullDumps | MiniDumps | ClusterManagementLog | ClusterLog | UpdateLogs | CbsLogs | StorageCmdlets | ClusterCmdlets | ConfigurationCmdlets | KernelDump | RollbackLogs | Symbols | NetworkCmdlets | NetworkCmds | Fltmc | ClusterStorageLogs | UTElement | UTFlag | SmbWmiProvider | TimeCmds | LocalUILogs | ClusterHealthLogs | BcdeditCommand | BitLockerCommand | DirStats | ComputeRolesLogs | ComputeCmdlets | DeviceGuard | Manifests | MeasuredBootLogs | Stats | PeriodicStatLogs | MigrationLogs | RollbackSupportPackage | ArchivedLogs | Default}] [-MinimumTimestamp <datetime>] [-MaximumTimestamp <datetime>] [-IncludeArchived] [-IncludePeriodicStats] [-Credential <pscredential>] [<CommonParameters>]The cmdlet collects logs from your device and copies those logs to a specified network or local share.
The parameters used are as follows:
-Path- Specify the network or the local path to copy support package to. (required)-Credential- Specify the credentials to access the protected path.-Zip- Specify to generate a zip file.-Include- Specify to include the components to be included in the support package. If not specified,Defaultis assumed.-IncludeArchived- Specify to include archived logs in the support package.-IncludePeriodicStats- Specify to include periodic stat logs in the support package.
View device information
Use the
Get-HcsApplianceInfoto get the information for your device.The following example shows the usage of this cmdlet:
[10.100.10.10]: PS>Get-HcsApplianceInfo Id : b2044bdb-56fd-4561-a90b-407b2a67bdfc FriendlyName : DBE-NBSVFQR94S6 Name : DBE-NBSVFQR94S6 SerialNumber : HCS-NBSVFQR94S6 DeviceId : 40d7288d-cd28-481d-a1ea-87ba9e71ca6b Model : Virtual FriendlySoftwareVersion : Data Box Gateway 1902 HcsVersion : 1.4.771.324 IsClustered : False IsVirtual : True LocalCapacityInMb : 1964992 SystemState : Initialized SystemStatus : Normal Type : DataBoxGateway CloudReadRateBytesPerSec : 0 CloudWriteRateBytesPerSec : 0 IsInitialPasswordSet : True FriendlySoftwareVersionNumber : 1902 UploadPolicy : All DataDiskResiliencySettingName : Simple ApplianceTypeFriendlyName : Data Box Gateway IsRegistered : FalseHere is a table summarizing some of the important device information:
Parameter Description FriendlyName The friendly name of the device as configured through the local web UI during device deployment. The default friendly name is the device serial number. SerialNumber The device serial number is a unique number assigned at the factory. Model The model for your Azure Stack Edge or Data Box Gateway device. The model is physical for Azure Stack Edge and virtual for Data Box Gateway. FriendlySoftwareVersion The friendly string that corresponds to the device software version. For a system running preview, the friendly software version would be Data Box Edge 1902. HcsVersion The HCS software version running on your device. For instance, the HCS software version corresponding to Data Box Edge 1902 is 1.4.771.324. LocalCapacityInMb The total local capacity of the device in Megabits. IsRegistered This value indicates if your device is activated with the service.
View GPU driver information
If the compute role is configured on your device, you can also get the GPU driver information via the PowerShell interface.
Use the
Get-HcsGpuNvidiaSmito get the GPU driver information for your device.The following example shows the usage of this cmdlet:
Get-HcsGpuNvidiaSmiMake a note of the driver information from the sample output of this cmdlet.
+-----------------------------------------------------------------------------+ | NVIDIA-SMI 440.64.00 Driver Version: 440.64.00 CUDA Version: 10.2 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 Tesla T4 On | 000029CE:00:00.0 Off | 0 | | N/A 60C P0 29W / 70W | 1539MiB / 15109MiB | 0% Default | +-------------------------------+----------------------+----------------------+ | 1 Tesla T4 On | 0000AD50:00:00.0 Off | 0 | | N/A 58C P0 29W / 70W | 330MiB / 15109MiB | 0% Default | +-------------------------------+----------------------+----------------------+
Enable Multi-Process Service (MPS)
A Multi-Process Service (MPS) on Nvidia GPUs provides a mechanism where GPUs can be shared by multiple jobs, where each job is allocated some percentage of the GPU's resources. MPS is a preview feature on your Azure Stack Edge Pro GPU device. To enable MPS on your device, follow these steps:
Before you begin, make sure that:
- You've configured and Activated your Azure Stack Edge Pro device with an Azure Stack Edge resource in Azure.
- You've Configured compute on this device in the Azure portal.
Use the following command to enable MPS on your device.
Start-HcsGpuMPS
Note
When the device software and the Kubernetes cluster are updated, the MPS setting is not retained for the workloads. You'll need to enable MPS again.
Reset your device
To reset your device, you need to securely wipe out all the data on the data disk and the boot disk of your device.
Use the Reset-HcsAppliance cmdlet to wipe out both the data disks and the boot disk or just the data disks. The SecureWipeBootDisk and SecureWipeDataDisks switches allow you to wipe the boot disk and the data disks respectively.
The SecureWipeBootDisk switch wipes the boot disk and makes the device unusable. It should be used only when the device needs to be returned to Microsoft. For more information, see Return the device to Microsoft.
If you use the device reset in the local web UI, only the data disks are securely wiped but the boot disk is kept intact. The boot disk contains the device configuration.
At the command prompt, type:
Reset-HcsAppliance -SecureWipeBootDisk -SecureWipeDataDisksThe following example shows how to use this cmdlet:
[10.128.24.33]: PS>Reset-HcsAppliance -SecureWipeBootDisk -SecureWipeDataDisks Confirm Are you sure you want to perform this action? Performing the operation "Reset-HcsAppliance" on target "ShouldProcess appliance". [Y] Yes [A] Yes to All [N] No [L] No to All [?] Help (default is "Y"): N
Get compute logs
If the compute role is configured on your device, you can also get the compute logs via the PowerShell interface.
Use the
Get-AzureDataBoxEdgeComputeRoleLogsto get the compute logs for your device.The following example shows the usage of this cmdlet:
Get-AzureDataBoxEdgeComputeRoleLogs -Path "\\hcsfs\logs\myacct" -Credential "username" -FullLogCollectionHere is a description of the parameters used for the cmdlet:
Path: Provide a network path to the share where you want to create the compute log package.Credential: Provide the username for the network share. When you run this cmdlet, you will need to provide the share password.FullLogCollection: This parameter ensures that the log package will contain all the compute logs. By default, the log package contains only a subset of logs.
Change Kubernetes workload profiles
After you have formed and configured a cluster and you have created new virtual switches, you can add or delete virtual networks associated with your virtual switches. For detailed steps, see Configure virtual switches.
After virtual switches are created, you can enable the switches for Kubernetes compute traffic to specify a Kubernetes workload profile. To do so using the local UI, use the steps in Configure compute IPS. To do so using PowerShell, use the following steps:
Use the
Get-HcsApplianceInfocmdlet to get currentKubernetesPlatformandKubernetesWorkloadProfilesettings for your device.Use the
Get-HcsKubernetesWorkloadProfilescmdlet to identify the profiles available on your Azure Stack Edge device.[Device-IP]: PS>Get-HcsKubernetesWorkloadProfiles Type Description ---- ----------- AP5GC an Azure Private MEC solution SAP a SAP Digital Manufacturing for Edge Computing or another Microsoft partner solution NONE other workloads [Device-IP]: PS>Use the
Set-HcsKubernetesWorkloadProfilecmdlet to set the workload profile for AP5GC, an Azure Private MEC solution.The following example shows the usage of this cmdlet:
Set-HcsKubernetesWorkloadProfile -Type "AP5GC"Here is sample output for this cmdlet:
[10.100.10.10]: PS>KubernetesPlatform : AKS [10.100.10.10]: PS>KubernetesWorkloadProfile : AP5GC [10.100.10.10]: PS>
Change Kubernetes pod and service subnets
If you're running the other workloads option in your environment, by default, Kubernetes on your Azure Stack Edge device uses subnets 172.27.0.0/16 and 172.28.0.0/16 for pod and service respectively. If these subnets are already in use in your network, then you can run the Set-HcsKubeClusterNetworkInfo cmdlet to change these subnets.
You want to perform this configuration before you configure compute from the Azure portal as the Kubernetes cluster is created in this step.
From the PowerShell interface of the device, run:
Set-HcsKubeClusterNetworkInfo -PodSubnet <subnet details> -ServiceSubnet <subnet details>Replace the <subnet details> with the subnet range that you want to use.
Once you have run this command, you can use the
Get-HcsKubeClusterNetworkInfocommand to verify that the pod and service subnets have changed.
Here is a sample output for this command.
[10.100.10.10]: PS>Set-HcsKubeClusterNetworkInfo -PodSubnet 10.96.0.1/16 -ServiceSubnet 10.97.0.1/16
[10.100.10.10]: PS>Get-HcsKubeClusterNetworkInfo
Id PodSubnet ServiceSubnet
-- --------- -------------
6dbf23c3-f146-4d57-bdfc-76cad714cfd1 10.96.0.1/16 10.97.0.1/16
[10.100.10.10]: PS>
Debug Kubernetes issues related to IoT Edge
Before you begin, you must have:
- Compute network configured. See Tutorial: Configure network for Azure Stack Edge Pro with GPU.
- Compute role configured on your device.
On an Azure Stack Edge Pro GPU device that has the compute role configured, you can troubleshoot or monitor the device using two different sets of commands.
- Using
iotedgecommands. These commands are available for basic operations for your device. - Using
kubectlcommands. These commands are available for an extensive set of operations for your device.
To execute either of the above set of commands, you need to Connect to the PowerShell interface.
Use iotedge commands
To see a list of available commands, connect to the PowerShell interface and use the iotedge function.
[10.100.10.10]: PS>iotedge -?
Usage: iotedge COMMAND
Commands:
list
logs
restart
[10.100.10.10]: PS>
The following table has a brief description of the commands available for iotedge:
| command | Description |
|---|---|
list |
List modules |
logs |
Fetch the logs of a module |
restart |
Stop and restart a module |
List all IoT Edge modules
To list all the modules running on your device, use the iotedge list command.
Here is a sample output of this command. This command lists all the modules, associated configuration, and the external IPs associated with the modules. For example, you can access the webserver app at https://10.128.44.244.
[10.100.10.10]: PS>iotedge list
NAME STATUS DESCRIPTION CONFIG EXTERNAL-IP
---- ------ ----------- ------ -----
gettingstartedwithgpus Running Up 10 days mcr.microsoft.com/intelligentedge/solutions:latest
iotedged Running Up 10 days azureiotedge/azureiotedge-iotedged:0.1.0-beta10 <none>
edgehub Running Up 10 days mcr.microsoft.com/azureiotedge-hub:1.0 10.128.44.243
edgeagent Running Up 10 days azureiotedge/azureiotedge-agent:0.1.0-beta10
webserverapp Running Up 10 days nginx:stable 10.128.44.244
[10.100.10.10]: PS>
Restart modules
You can use the list command to list all the modules running on your device. Then identify the name of the module that you want to restart and use it with the restart command.
Here is a sample output of how to restart a module. Based on the description of how long the module is running for, you can see that cuda-sample1 was restarted.
[10.100.10.10]: PS>iotedge list
NAME STATUS DESCRIPTION CONFIG EXTERNAL-IP PORT(S)
---- ------ ----------- ------ ----------- -------
edgehub Running Up 5 days mcr.microsoft.com/azureiotedge-hub:1.0 10.57.48.62 443:31457/TCP,5671:308
81/TCP,8883:31753/TCP
iotedged Running Up 7 days azureiotedge/azureiotedge-iotedged:0.1.0-beta13 <none> 35000/TCP,35001/TCP
cuda-sample2 Running Up 1 days nvidia/samples:nbody
edgeagent Running Up 7 days azureiotedge/azureiotedge-agent:0.1.0-beta13
cuda-sample1 Running Up 1 days nvidia/samples:nbody
[10.100.10.10]: PS>iotedge restart cuda-sample1
[10.100.10.10]: PS>iotedge list
NAME STATUS DESCRIPTION CONFIG EXTERNAL-IP PORT(S)
---- ------ ----------- ------ ----------- -------
edgehub Running Up 5 days mcr.microsoft.com/azureiotedge-hub:1.0 10.57.48.62 443:31457/TCP,5671:30
881/TCP,8883:31753/TC
P
iotedged Running Up 7 days azureiotedge/azureiotedge-iotedged:0.1.0-beta13 <none> 35000/TCP,35001/TCP
cuda-sample2 Running Up 1 days nvidia/samples:nbody
edgeagent Running Up 7 days azureiotedge/azureiotedge-agent:0.1.0-beta13
cuda-sample1 Running Up 4 minutes nvidia/samples:nbody
[10.100.10.10]: PS>
Get module logs
Use the logs command to get logs for any IoT Edge module running on your device.
If there was an error in creation of the container image or while pulling the image, run logs edgeagent. edgeagent is the IoT Edge runtime container that is responsible for provisioning other containers. Because logs edgeagent dumps all the logs, a good way to see the recent errors is to use the option --tail 0`.
Here is a sample output.
[10.100.10.10]: PS>iotedge logs cuda-sample2 --tail 10
[10.100.10.10]: PS>iotedge logs edgeagent --tail 10
<6> 2021-02-25 00:52:54.828 +00:00 [INF] - Executing command: "Report EdgeDeployment status: [Success]"
<6> 2021-02-25 00:52:54.829 +00:00 [INF] - Plan execution ended for deployment 11
<6> 2021-02-25 00:53:00.191 +00:00 [INF] - Plan execution started for deployment 11
<6> 2021-02-25 00:53:00.191 +00:00 [INF] - Executing command: "Create an EdgeDeployment with modules: [cuda-sample2, edgeAgent, edgeHub, cuda-sample1]"
<6> 2021-02-25 00:53:00.212 +00:00 [INF] - Executing command: "Report EdgeDeployment status: [Success]"
<6> 2021-02-25 00:53:00.212 +00:00 [INF] - Plan execution ended for deployment 11
<6> 2021-02-25 00:53:05.319 +00:00 [INF] - Plan execution started for deployment 11
<6> 2021-02-25 00:53:05.319 +00:00 [INF] - Executing command: "Create an EdgeDeployment with modules: [cuda-sample2, edgeAgent, edgeHub, cuda-sample1]"
<6> 2021-02-25 00:53:05.412 +00:00 [INF] - Executing command: "Report EdgeDeployment status: [Success]"
<6> 2021-02-25 00:53:05.412 +00:00 [INF] - Plan execution ended for deployment 11
[10.100.10.10]: PS>
Note
The direct methods such as GetModuleLogs or UploadModuleLogs are not supported on IoT Edge on Kubernetes on your Azure Stack Edge.
Use kubectl commands
On an Azure Stack Edge Pro GPU device that has the compute role configured, all the kubectl commands are available to monitor or troubleshoot modules. To see a list of available commands, run kubectl --help from the command window.
C:\Users\myuser>kubectl --help
kubectl controls the Kubernetes cluster manager.
Find more information at: https://kubernetes.io/docs/reference/kubectl/overview/
Basic Commands (Beginner):
create Create a resource from a file or from stdin.
expose Take a replication controller, service, deployment or pod and expose it as a new Kubernetes Service
run Run a particular image on the cluster
set Set specific features on objects
run-container Run a particular image on the cluster. This command is deprecated, use "run" instead
==============CUT=============CUT============CUT========================
Usage:
kubectl [flags] [options]
Use "kubectl <command> --help" for more information about a given command.
Use "kubectl options" for a list of global command-line options (applies to all commands).
C:\Users\myuser>
For a comprehensive list of the kubectl commands, go to kubectl cheatsheet.
To get IP of service or module exposed outside of Kubernetes cluster
To get the IP of a load-balancing service or modules exposed outside of the Kubernetes, run the following command:
kubectl get svc -n iotedge
Following is a sample output of the all the services or modules that are exposed outside of the Kubernetes cluster.
[10.100.10.10]: PS>kubectl get svc -n iotedge
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
edgehub LoadBalancer 10.103.52.225 10.128.44.243 443:31987/TCP,5671:32336/TCP,8883:30618/TCP 34h
iotedged ClusterIP 10.107.236.20 <none> 35000/TCP,35001/TCP 3d8h
webserverapp LoadBalancer 10.105.186.35 10.128.44.244 8080:30976/TCP 16h
[10.100.10.10]: PS>
The IP address in the External IP column corresponds to the external endpoint for the service or the module. You can also Get the external IP in the Kubernetes dashboard.
To check if module deployed successfully
Compute modules are containers that have a business logic implemented. A Kubernetes pod can have multiple containers running.
To check if a compute module is deployed successfully, connect to the PowerShell interface of the device.
Run the get pods command and check if the container (corresponding to the compute module) is running.
To get the list of all the pods running in a specific namespace, run the following command:
get pods -n <namespace>
To check the modules deployed via IoT Edge, run the following command:
get pods -n iotedge
Following is a sample output of all the pods running in the iotedge namespace.
[10.100.10.10]: PS>kubectl get pods -n iotedge
NAME READY STATUS RESTARTS AGE
edgeagent-cf6d4ffd4-q5l2k 2/2 Running 0 20h
edgehub-8c9dc8788-2mvwv 2/2 Running 0 56m
filemove-66c49984b7-h8lxc 2/2 Running 0 56m
iotedged-675d7f4b5f-9nml4 1/1 Running 0 20h
[10.100.10.10]: PS>
The status Status indicates that all the pods in the namespace are running and the Ready indicates the number of containers deployed in a pod. In the preceding sample, all the pods are running and all the modules deployed in each of the pods are running.
To check the modules deployed via Azure Arc, run the following command:
get pods -n azure-arc
Alternatively, you can Connect to Kubernetes dashboard to see IoT Edge or Azure Arc deployments.
For a more verbose output of a specific pod for a given namespace, you can run the following command:
kubectl describe pod <pod name> -n <namespace>
The sample output is shown here.
[10.100.10.10]: PS>kubectl describe pod filemove-66c49984b7 -n iotedge
Name: filemove-66c49984b7-h8lxc
Namespace: iotedge
Priority: 0
Node: k8s-1hwf613cl-1hwf613/10.139.218.12
Start Time: Thu, 14 May 2020 12:46:28 -0700
Labels: net.azure-devices.edge.deviceid=myasegpu-edge
net.azure-devices.edge.hub=myasegpu2iothub.azure-devices.net
net.azure-devices.edge.module=filemove
pod-template-hash=66c49984b7
Annotations: net.azure-devices.edge.original-moduleid: filemove
Status: Running
IP: 172.17.75.81
IPs: <none>
Controlled By: ReplicaSet/filemove-66c49984b7
Containers:
proxy:
Container ID: docker://fd7975ca78209a633a1f314631042a0892a833b7e942db2e7708b41f03e8daaf
Image: azureiotedge/azureiotedge-proxy:0.1.0-beta8
Image ID: docker://sha256:5efbf6238f13d24bab9a2b499e5e05bc0c33ab1587d6cf6f289cdbe7aa667563
Port: <none>
Host Port: <none>
State: Running
Started: Thu, 14 May 2020 12:46:30 -0700
Ready: True
Restart Count: 0
Environment:
PROXY_LOG: Debug
=============CUT===============================CUT===========================
Volumes:
config-volume:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: iotedged-proxy-config
Optional: false
trust-bundle-volume:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: iotedged-proxy-trust-bundle
Optional: false
myasesmb1local:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: myasesmb1local
ReadOnly: false
myasesmb1:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: myasesmb1
ReadOnly: false
filemove-token-pzvw8:
Type: Secret (a volume populated by a Secret)
SecretName: filemove-token-pzvw8
Optional: false
QoS Class: BestEffort
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events: <none>
[10.100.10.10]: PS>
To get container logs
To get the logs for a module, run the following command from the PowerShell interface of the device:
kubectl logs <pod_name> -n <namespace> --all-containers
Because all-containers flag dumps all the logs for all the containers, a good way to see the recent errors is to use the option --tail 10.
Following is a sample output.
[10.100.10.10]: PS>kubectl logs filemove-66c49984b7-h8lxc -n iotedge --all-containers --tail 10
DEBUG 2020-05-14T20:40:42Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:40:44Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:40:44Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:40:44Z: loop process - 1 events, 0.000s
DEBUG 2020-05-14T20:40:44Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:42:12Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:42:14Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:42:14Z: loop process - 0 events, 0.000s
DEBUG 2020-05-14T20:42:14Z: loop process - 1 events, 0.000s
DEBUG 2020-05-14T20:42:14Z: loop process - 0 events, 0.000s
05/14/2020 19:46:44: Info: Opening module client connection.
05/14/2020 19:46:45: Info: Open done.
05/14/2020 19:46:45: Info: Initializing with input: /home/input, output: /home/output, protocol: Amqp.
05/14/2020 19:46:45: Info: IoT Hub module client initialized.
[10.100.10.10]: PS>
Change memory, processor limits for Kubernetes worker node
To change the memory or processor limits for Kubernetes worker node, do the following steps:
To get the current resources for the worker node and the role options, run the following command:
Get-AzureDataBoxEdgeRoleHere is a sample output. Note the values for
NameandComputeunderResourcessection.MemoryInBytesandProcessorCountdenote the currently assigned values memory and processor count for the Kubernetes worker node.[10.100.10.10]: PS>Get-AzureDataBoxEdgeRole ImageDetail : Name:mcr.microsoft.com/azureiotedge-agent Tag:1.0 PlatformType:Linux EdgeDeviceConnectionString : IotDeviceConnectionString : HubHostName : ase-srp-007.azure-devices.net IotDeviceId : srp-007-storagegateway EdgeDeviceId : srp-007-edge Version : Id : 6ebeff9f-84c5-49a7-890c-f5e05520a506 Name : IotRole Type : IOT Resources : Compute: MemoryInBytes:34359738368 ProcessorCount:12 VMProfile: Storage: EndpointMap: EndpointId:c0721210-23c2-4d16-bca6-c80e171a0781 TargetPath:mysmbedgecloudshare1 Name:mysmbedgecloudshare1 Protocol:SMB EndpointId:6557c3b6-d3c5-4f94-aaa0-6b7313ab5c74 TargetPath:mysmbedgelocalshare Name:mysmbedgelocalshare Protocol:SMB RootFileSystemStorageSizeInBytes:0 HostPlatform : KubernetesCluster State : Created PlatformType : Linux HostPlatformInstanceId : 994632cb-853e-41c5-a9cd-05b36ddbb190 IsHostPlatformOwner : True IsCreated : True [10.100.10.10]: PS>To change the values of memory and processors for the worker node, run the following command:
Set-AzureDataBoxEdgeRoleCompute -Name <Name value from the output of Get-AzureDataBoxEdgeRole> -Memory <Value in Bytes> -ProcessorCount <No. of cores>Here is a sample output.
[10.100.10.10]: PS>Set-AzureDataBoxEdgeRoleCompute -Name IotRole -MemoryInBytes 32GB -ProcessorCount 16 ImageDetail : Name:mcr.microsoft.com/azureiotedge-agent Tag:1.0 PlatformType:Linux EdgeDeviceConnectionString : IotDeviceConnectionString : HubHostName : ase-srp-007.azure-devices.net IotDeviceId : srp-007-storagegateway EdgeDeviceId : srp-007-edge Version : Id : 6ebeff9f-84c5-49a7-890c-f5e05520a506 Name : IotRole Type : IOT Resources : Compute: MemoryInBytes:34359738368 ProcessorCount:16 VMProfile: Storage: EndpointMap: EndpointId:c0721210-23c2-4d16-bca6-c80e171a0781 TargetPath:mysmbedgecloudshare1 Name:mysmbedgecloudshare1 Protocol:SMB EndpointId:6557c3b6-d3c5-4f94-aaa0-6b7313ab5c74 TargetPath:mysmbedgelocalshare Name:mysmbedgelocalshare Protocol:SMB RootFileSystemStorageSizeInBytes:0 HostPlatform : KubernetesCluster State : Created PlatformType : Linux HostPlatformInstanceId : 994632cb-853e-41c5-a9cd-05b36ddbb190 IsHostPlatformOwner : True IsCreated : True [10.100.10.10]: PS>
While changing the memory and processor usage, follow these guidelines.
- Default memory is 25% of device specification.
- Default processor count is 30% of device specification.
- When changing the values for memory and processor counts, we recommend that you vary the values between 15% to 60% of the device memory and the processor count.
- We recommend an upper limit of 60% is so that there are enough resources for system components.
Connect to BMC
Note
Baseboard management controller (BMC) is not available on Azure Stack Edge Pro 2 and Azure Stack Edge Mini R. The cmdlets described in this section only apply to Azure Stack Edge Pro GPU and Azure Stack Edge Pro R.
BMC is used to remotely monitor and manage your device. This section describes the cmdlets that can be used to manage BMC configuration. Prior to running any of these cmdlets, Connect to the PowerShell interface of the device.
Get-HcsNetBmcInterface: Use this cmdlet to get the network configuration properties of the BMC, for example,IPv4Address,IPv4Gateway,IPv4SubnetMask,DhcpEnabled.Here is a sample output:
[10.100.10.10]: PS>Get-HcsNetBmcInterface IPv4Address IPv4Gateway IPv4SubnetMask DhcpEnabled ----------- ----------- -------------- ----------- 10.128.53.186 10.128.52.1 255.255.252.0 False [10.100.10.10]: PS>Set-HcsNetBmcInterface: You can use this cmdlet in the following two ways.Use the cmdlet to enable or disable DHCP configuration for BMC by using the appropriate value for
UseDhcpparameter.Set-HcsNetBmcInterface -UseDhcp $trueHere is a sample output:
[10.100.10.10]: PS>Set-HcsNetBmcInterface -UseDhcp $true [10.100.10.10]: PS>Get-HcsNetBmcInterface IPv4Address IPv4Gateway IPv4SubnetMask DhcpEnabled ----------- ----------- -------------- ----------- 10.128.54.8 10.128.52.1 255.255.252.0 True [10.100.10.10]: PS>Use this cmdlet to configure the static configuration for the BMC. You can specify the values for
IPv4Address,IPv4Gateway, andIPv4SubnetMask.Set-HcsNetBmcInterface -IPv4Address "<IPv4 address of the device>" -IPv4Gateway "<IPv4 address of the gateway>" -IPv4SubnetMask "<IPv4 address for the subnet mask>"Here is a sample output:
[10.100.10.10]: PS>Set-HcsNetBmcInterface -IPv4Address 10.128.53.186 -IPv4Gateway 10.128.52.1 -IPv4SubnetMask 255.255.252.0 [10.100.10.10]: PS>Get-HcsNetBmcInterface IPv4Address IPv4Gateway IPv4SubnetMask DhcpEnabled ----------- ----------- -------------- ----------- 10.128.53.186 10.128.52.1 255.255.252.0 False [10.100.10.10]: PS>
Set-HcsBmcPassword: Use this cmdlet to modify the BMC password forEdgeUser. The user name -EdgeUser- is case-sensitive.Here is a sample output:
[10.100.10.10]: PS> Set-HcsBmcPassword -NewPassword "Password1" [10.100.10.10]: PS>
Exit the remote session
To exit the remote PowerShell session, close the PowerShell window.
Next steps
- Deploy Azure Stack Edge Pro GPU in Azure portal.
Feedback
Coming soon: Throughout 2024 we will be phasing out GitHub Issues as the feedback mechanism for content and replacing it with a new feedback system. For more information see: https://aka.ms/ContentUserFeedback.
Submit and view feedback for