Azure
Last updated
Was this helpful?
XMPro Platform
Last updated
Was this helpful?
This document will guide you through how to set up the Azure infrastructure and deploy the XMPro Platform.
The following deployment diagram shows an example architecture and the necessary resources for the XMPro platform in Microsoft Azure.
Set up of Cloud Stream Host, Master Data, Azure Cache, or Twilio is optional and is dependent on client requirements. As a result, these resources will be excluded from this deployment.
In order to proceed with the deployment, you are required to complete the steps in the 1. Preparation guide:
Select Azure resources that meet the hardware requirements and software requirements.
Follow the certificate and communication steps for an SMTP Account (Recommended).
As well as having:
Azure portal administrative access
Access to Subscription and Resource group
This section provides information about installing the XMPro platform from scratch in Azure.
Log on to the Azure Portal https://portal.azure.com/ (with Company Administrator access)
In the search bar type “Deploy” and select “Deploy a custom template”
3. Click "Build your own template in the editor"
4. Download the file from the link provided below:
5. Select "Load File" and then Open the downloaded "mainTemplate_[Version].json"
6. Once loaded, click Save - do not change the template.
7. Complete the form.
Enter the username in the correct format i.e. firstname.lastname@companyname.onxmpro.com
The following special characters are not supported in the passwords: ` ' \ $ @
Take note of the passwords used as this will not be displayed again and they are required later.
8. Verify the information is correct and click Create.
After successful deployment, the following items have been installed:
Subscription Manager
Data Stream Designer
App Designer
Stream Host
Proceed to Restart the App Services after they are all deployed.
This step explains how to restart your app services in the Azure Portal.
Type “Resource groups” in the search bar and select “Resource groups”.
2. Search for the Resource Group created during installation and select it.
3. Select Subscription Manager, Data Stream Designer, and App Designer and restart the applications.
4. The below GIF shows how to restart your app service.
admin@xmpro.onxmpro.com
Super Admin
as entered during setup
firstname.lastname@companyname.onxmpro.com
Admin
as entered during setup
A new company is created as part of the installation process but needs a valid license to work.
1. Click Company in the left menu to open the Companies page. 2. Click on the Company. 3. Click on the Subscriptions gauge to open the Subscriptions page. 4. Click on a Subscription. 5. Click the Update License button in the command bar. 6. Click Generate a license request, enter the number of days, and submit. 7. When you have received the license from XMPro support, upload it. 8. Click Save.
This sends a request to XMPro for an App Designer license for this Company.
The Generate a license request link will only work if SMTP was set up during installation.
9. Change the Product to Data Stream Designer from the list and request another license
When you have received a license for each product through an email sent to the email address given during installation, follow the steps below to upload the license for both App Designer and Data Stream Designer.
The installation of the XMPro Platform is now complete but before you can use the platform, some steps are needed to set up the environment. Further instructions about the configuration can be found below:
Before beginning the upgrade, back up databases using this guide.
Follow the same steps as a new Install to upgrade your XMPro platform**.** Make sure to:
Download and use the latest "mainTemplate_[Version].json".
Before deploying, stop the SM App Service to avoid file locking issues during MSDeploy.
Use the same Resource Group and credentials used during the original Install when completing the Custom deployment form.
When upgrading XMPro instances that use AI Notebooks installed via ARM deployments, you may encounter known issues with the automated upgrade process. These issues primarily affect the Notebook components during ARM-based upgrades.
This issue affects users who deploy via ARM and already have AI Notebooks installed
The issue primarily appears during upgrades to version 4.4.18 and later
The problem involves compatibility between the ARM template and current Azure implementation
While automated Notebook upgrades via ARM deployments may fail, you can successfully upgrade Notebooks using the following manual deployment process:
Bash terminal
Azure CLI 2
Helm 3.X
Login and validate connection to the correct subscription
az login
az account showFind the name of the AKS resource
az aks list -o tableMerge the AKS cluster into kubeconfig
az aks get-credentials -n <aks resource name> -g <resource group>Check the kubectl is connected to the AKS instance
kubectl config current-contextFind the helm deployment
helm list -n prodCheck the helm deployment values and save to values.yml
helm get values xmpro-notebooks -n prod > values.yml | sed -i '1d' values.ymlRemove the first line "USER-SUPPLIED VALUES:"
sed -i '1d' values.ymlObserve the singleuser section in values.yml
singleuser:
image:
name: <registry>/base-jupyter-notebook
tag: <release-version>Set the following values
<registry> = xmpro.azurecr.io
<release-version> = the desired release tag version e.g. 4.4.18
Deploy the values upgrade using Helm
helm upgrade --cleanup-on-fail --install xmpro-notebooks jupyterhub/jupyterhub --namespace prod --create-namespaceWait for the deployment to complete
Verify the deployment
helm list -n prodVerify the values are set correctly
helm get values xmpro-notebooks -n prodVerify the pods status
kubectl get pods -n prodIf required, forcefully restart the notebook pods
kubectl delete pods --all -n prodThe URLs for notebook should be in sync in all the places:
AI AppService > Environment Vars > xmpro:aidesigner:jupyterUrl
SM DB dbo.ProductUrls > Notebook Url
Helm Chart Release > xmpro-notebook > values
kubectl describe svc proxy-public -n prod > annotations > service.beta.kubernetes.io/azure-dns-label-name
// az login to the subscriptionConnect to the aks instance
az aks list -o table
az aks get-credentials -g <resource-group> -n <aks-instance-name-from-above>
kubectl config set-context <aks-instance-name-from-above>
kubectl config current-contextGet helm deployment details
helm ls -n prod
helm get values xmpro-notebooks -n prod
// check the notebook URLsGet proxy public annotations
kubectl get svc -n prod
kubectl describe svc proxy-public -n prod
// check the service.beta.kubernetes.io/azure-dns-label-name annotationCheck the SQL database
az resource list -g <resource-group> -o table
az sql server show -g <resource-group> -n <sql-server-instance-name-from-above>
// check the SM DB dbo.ProductUrls > Notebook UrlForce restart all pods
kubectl get pods -n prod
kubectl delete pods --all -n prodAfter a Notebook upgrade, launching a Jupyter Hub user instance may result in an error such as;
MountVolume.MountDevice failed for volume "pvc-64bd7714-7c0e-45ca-88cc-c1bdf816108e": rpc error: code = Internal desc = volume(MC_armupgrade_aks-armup-xmpro-001_centralus#nfs0f17de3935cc49c2ae46#pvc-64bd7714-7c0e-45ca-88cc-c1bdf816108e##prod#) mount "nfs0f17de3935cc49c2ae46.blob.core.windows.net:/nfs0f17de3935cc49c2ae46/pvc-64bd7714-7c0e-45ca-88cc-c1bdf816108e" on "/var/lib/kubelet/plugins/kubernetes.io/csi/blob.csi.azure.com/fd5793275cbf92e1d702e30d03bcf1634564e3690e3b635a763e814f5e09c29/globalmount" failed with mount failed: exit status 1 Mounting command: mount Mounting arguments: -t aznfs -o nconnect=4,sec=sys,vers=3,nolock nfs0f17de3935cc49c2ae46.blob.core.windows.net:/nfs0f17de3935cc49c2ae46/pvc-64bd7714-7c0e-45ca-88cc-c1bdf816108e /var/lib/kubelet/plugins/kubernetes.io/csi/blob.csi.azure.com/fd5793275cbf92e1d702e30d03bcf1634564e3690e3b635a763e814f5e09c29/globalmount Output: [0mmount.nfs: access denied by server while mounting 10.161.100.100:/nfs0f17de3935cc49c2ae46/pvc-64bd7714-7c0e-45ca-88cc-c1bdf816108e[0m [2;31mMount failed
