Trying to update an email account to new IMAP and SMTP details and you can’t do that because that feature does not exist, the only option is to delete the account and start over.

However, there are emails only on the machine in outlook for the account that we need to keep, which wouldn’t be a problem in the old outlook, we could just export them…but not with new outlook, now we can’t export files because once again that feature does not exist.

Why would they release such an incomplete product and tell people it’s new and improved?

Ended up downloading a different mail app for the new email address just so we could keep the old emails, such a farce.

​Trying to update an email account to new IMAP and SMTP details and you can’t do that because that feature does not exist, the only option is to delete the account and start over.However, there are emails only on the machine in outlook for the account that we need to keep, which wouldn’t be a problem in the old outlook, we could just export them…but not with new outlook, now we can’t export files because once again that feature does not exist.Why would they release such an incomplete product and tell people it’s new and improved?Ended up downloading a different mail app for the new email address just so we could keep the old emails, such a farce.  Read More

​

My customer is using MRM and has been for a long time in office 365.  Recently they noticed that employees were putting personal tags on some items, this is against their policy.  The way to prevent users from using personal tags is documented here.
Users can use all personal retention tags regardless of retention policy in Exchange Online – Microsoft Support

Once that is complete, they’d like to remove any personal tags on email items or folders in all user mailboxes.  I am told you need an EWS script to do this.  Does anyone have an examples script?

​My customer is using MRM and has been for a long time in office 365.  Recently they noticed that employees were putting personal tags on some items, this is against their policy.  The way to prevent users from using personal tags is documented here.Users can use all personal retention tags regardless of retention policy in Exchange Online – Microsoft SupportOnce that is complete, they’d like to remove any personal tags on email items or folders in all user mailboxes.  I am told you need an EWS script to do this.  Does anyone have an examples script?  Read More

​

Dear Community,

I am desperately trying my luck here in this community. Can anyone help me how to highlight lowest value in multiple excel sheet?

Thank you in advance for your help

​Dear Community,I am desperately trying my luck here in this community. Can anyone help me how to highlight lowest value in multiple excel sheet?Thank you in advance for your help  Read More

​

I am attempting to build an Azure Virtual Desktop that will be used for an Exam Environment. The physical workstations are thin clients that auto-direct to RDP and access our AVD environment. Using a Windows11+Office365 image, I created a basic image. I added shortcuts to the desktop (Word, Excel, PowerPoint, a web shortcut)

I want a non-persistent desktop that does not pull any information from a user’s account. Not OneDrive, not Teams, no roaming files. It will pass-through the user’s credentials for obtaining a Office license and nothing more. All internet will be blocked, and only a couple of sites will be accessed.

All lockdown settings are handled through GPO (GPO settings File). **not shown in the file is the usage of Teams where I used a fake tenant ID so it was impossible to loadlink MSTeams.  But to block OneDrive as indicated in the file, I did use the correct tenant ID.

Does it work? Yes.  But it has issues loading up. 
When you attempt to sign in, it will take exactly 10min ‘Preparing Windows’.
Once this completes, in the bottom left corner is a minimized DOS window. When you restore this window it shows it’s attempting to run UsrLogon.cmd but has the error “The command prompt has been disabled by your administrator. Press any key to continue…”, which when done takes you to the desktop.

Lastly, the weblink is created for the student to upload their exam to our CRM when completed. When you access this weblink it opens with “Welcome-new-device is blocked” and you have to use TaskManager to kill Edge, and then restart it so it will then properly access the CRM site.

While this does work as intended once you get to this point, it’s an unnecessary delay. Especially when attempting to take an exam.

I need help trying to identify those (3) problems;

– Why is it taking 10 minutes to prepare windows, and how do you make it stop doing that?

– Why is the usrlogon.cmd failing to run (and I don’t appear to need it, so how to stop it completely?)

– What GPO did I miss with Edge to not have that ‘welcome new deviceuser’ to appear and just load the website I want.

​I am attempting to build an Azure Virtual Desktop that will be used for an Exam Environment. The physical workstations are thin clients that auto-direct to RDP and access our AVD environment. Using a Windows11+Office365 image, I created a basic image. I added shortcuts to the desktop (Word, Excel, PowerPoint, a web shortcut)I want a non-persistent desktop that does not pull any information from a user’s account. Not OneDrive, not Teams, no roaming files. It will pass-through the user’s credentials for obtaining a Office license and nothing more. All internet will be blocked, and only a couple of sites will be accessed.All lockdown settings are handled through GPO (GPO settings File). **not shown in the file is the usage of Teams where I used a fake tenant ID so it was impossible to loadlink MSTeams.  But to block OneDrive as indicated in the file, I did use the correct tenant ID. Does it work? Yes.  But it has issues loading up. When you attempt to sign in, it will take exactly 10min ‘Preparing Windows’.Once this completes, in the bottom left corner is a minimized DOS window. When you restore this window it shows it’s attempting to run UsrLogon.cmd but has the error “The command prompt has been disabled by your administrator. Press any key to continue…”, which when done takes you to the desktop.Lastly, the weblink is created for the student to upload their exam to our CRM when completed. When you access this weblink it opens with “Welcome-new-device is blocked” and you have to use TaskManager to kill Edge, and then restart it so it will then properly access the CRM site. While this does work as intended once you get to this point, it’s an unnecessary delay. Especially when attempting to take an exam. I need help trying to identify those (3) problems;- Why is it taking 10 minutes to prepare windows, and how do you make it stop doing that?- Why is the usrlogon.cmd failing to run (and I don’t appear to need it, so how to stop it completely?)- What GPO did I miss with Edge to not have that ‘welcome new deviceuser’ to appear and just load the website I want.  Read More

​

Hi all – Strange problem.  I have inherited a system where a user fills out a form and, on submission, a subsite is generated (Classic Mode).  About a month ago, “Edit Page” ceased working.  All it does is reload the page.  “Stop Editing” is greyed out and no web parts are selectable.  I viewed in Devleoper Mode (Edge) and there are no errors or JS bugs.  Has Microsoft deprecated features in Classic Mode that would affect editing web parts in Classic?

​Hi all – Strange problem.  I have inherited a system where a user fills out a form and, on submission, a subsite is generated (Classic Mode).  About a month ago, “Edit Page” ceased working.  All it does is reload the page.  “Stop Editing” is greyed out and no web parts are selectable.  I viewed in Devleoper Mode (Edge) and there are no errors or JS bugs.  Has Microsoft deprecated features in Classic Mode that would affect editing web parts in Classic?  Read More

​

Good evening!

I manage a fleet of laptops using Microsoft Intune.
Our Microsoft 365 Business Premium subscriptions include Sharepoint Online and I’d like to use it as a file server.
Everything works fine except for one specific point: locking files during editing.
With a Windows Server file share, when someone wants to access a file being edited, a pop-up appears, inviting the user to open it in read-only mode, notify the editor, etc., which avoids the risk of conflict.

With Sharepoint, it’s more complicated.

Here’s the scenario:

Sharepoint files are synchronized on PCs via OneDrive with the Intune configuration “Configure team site libraries to sync automatically, Use OneDrive Files on-demand and Convert synced team site files to online-only files” enabled.User 1 double-clicks on one of the files in the OneDrive folder, and a local copy is downloaded.User 1 edits the file via Word installed locally, while the Internet connection is lost.At the same time, user 2 edits the file and saves it. His version ends up on Sharepoint.A few minutes later, once User 1’s connection is re-established, his version is saved on Sharepoint.
User 2’s changes will go completely unnoticed, creating confusion and wasted time.

The check-out/check-in system proposed by Sharepoint has the merit of existing, but is very restrictive for users.

Is there a way of ensuring that Sharepoint files appear in the OneDrive folder in Windows File Explorer, but that modifications can only be made online, to avoid this type of problem?

I thought that the “Convert synced team site files to online-only files” option would meet this need, but it seems that this is not the case.

Thanks for your help!
Jo

​Good evening! I manage a fleet of laptops using Microsoft Intune.Our Microsoft 365 Business Premium subscriptions include Sharepoint Online and I’d like to use it as a file server.Everything works fine except for one specific point: locking files during editing.With a Windows Server file share, when someone wants to access a file being edited, a pop-up appears, inviting the user to open it in read-only mode, notify the editor, etc., which avoids the risk of conflict.With Sharepoint, it’s more complicated.Here’s the scenario:Sharepoint files are synchronized on PCs via OneDrive with the Intune configuration “Configure team site libraries to sync automatically, Use OneDrive Files on-demand and Convert synced team site files to online-only files” enabled.User 1 double-clicks on one of the files in the OneDrive folder, and a local copy is downloaded.User 1 edits the file via Word installed locally, while the Internet connection is lost.At the same time, user 2 edits the file and saves it. His version ends up on Sharepoint.A few minutes later, once User 1’s connection is re-established, his version is saved on Sharepoint.User 2’s changes will go completely unnoticed, creating confusion and wasted time.The check-out/check-in system proposed by Sharepoint has the merit of existing, but is very restrictive for users. Is there a way of ensuring that Sharepoint files appear in the OneDrive folder in Windows File Explorer, but that modifications can only be made online, to avoid this type of problem? I thought that the “Convert synced team site files to online-only files” option would meet this need, but it seems that this is not the case. Thanks for your help!Jo  Read More

​

I’m setting up a SharePoint list to support attestation by owners for different business virtual assets.  This will support a review to ensure we aren’t retaining stale virtual assets. 

When the owners review SharePoint list rows where they are listed as an owner, I’d like them to provide a note detailing their rationale for why the asset should be retained but have that note timestamped so the next time they perform a review they can see the previous rationale and the date they added that note. 

What is the best way to do this? Calculated, multi-line column?  If so, what formula would work? Or would this need to be run through Power Automate and have a flow that updates any input to that column to concatenate it with a date?

​I’m setting up a SharePoint list to support attestation by owners for different business virtual assets.  This will support a review to ensure we aren’t retaining stale virtual assets.  When the owners review SharePoint list rows where they are listed as an owner, I’d like them to provide a note detailing their rationale for why the asset should be retained but have that note timestamped so the next time they perform a review they can see the previous rationale and the date they added that note.  What is the best way to do this? Calculated, multi-line column?  If so, what formula would work? Or would this need to be run through Power Automate and have a flow that updates any input to that column to concatenate it with a date?  Read More

​

In the first part of this topic, we setup all the Azure resources like the Azure Communication Services for Email, SMS, WhatsApp for Business and developed the Azure Functions code for the email trigger. In this second part, we will complete coding the remaining Azure Functions triggers and then go ahead to deploy the multichannel notification system to Azure Functions, testing the Email, SMS, and WhatsApp triggers with OpenAI GPTs. Let’s get started!

Prerequisite​

To follow this tutorial, ensure you have completed the first part of this topic.

Coding the SMSTrigger​

Enhancing the SMSTrigger Azure Function from the default template involves a series of steps. These steps will transform the basic Function into one that can send SMS messages using Azure Communication Services. Below is a guide to get you from the default HTTP triggered function to the finished SMSTrigger.

Step 1: Set Up the Function Template​

Follow the instructions for setting up the function template from the Email section and name the trigger as ‘SMSTrigger’ or any other string you prefer.

Step 2: Add Azure Communication Services SMS Reference​

Add a reference to using Azure.Communication.Sms then create a property in the SMS Trigger class to hold an instance of SmsClient and a property to hold the email sender address.

csharp

Step 3: Read Configuration and Initialize SmsClient​

In the constructor of the SMSTrigger class, read the Azure Communication Services connection string from the environment variables using the Environment.GetEnvironmentVariable() method and initialize the SmsClient instance.

Be sure to check if the connection string is null, and if so, throw an exception to indicate that the environment variable is missing:

csharp

Step 4: Define the Request Model​

Create a request model class within the SMSTrigger class called SmsRequest. This model should contain properties for the message text and the phone number to which the message will be sent.

csharp

Step 5: Parse the Request Body​

Change the Run function to be async as we will perform asynchronous operations. Use a StreamReader to read the request body as a string and deserialize it into an SmsRequest object using JsonSerializer.

csharp

If the request body fails to deserialize into SmsRequest, return a BadRequestResult:

csharp

Step 6: Define the Sender and Send an SMS​

Retrieve the sender’s phone number from the environment variables with Environment.GetEnvironmentVariable(). Then, attempt to send the SMS with a try-catch block, handling any RequestFailedException that may occur and logging the relevant information:

csharp

Step 7: Return a Success Response​

If sending the SMS is successful, return an OkObjectResult to the caller indicating that the SMS has been sent.

csharp

Final Code​

The final SMSTrigger Azure Function, with the steps implemented, should look as follows:

csharp

This completed SMSTrigger Azure Function can now facilitate SMS as part of your multichannel notification system.

Coding the WhatsAppTrigger​

Creating a functional WhatsAppTrigger Azure Function involves iterating on the default HTTP-triggered function template provided by Azure Functions for C#. We will modify this template to integrate Azure Communication Services for sending WhatsApp messages via template messages. Follow the steps below to transform this template into a complete WhatsAppTrigger function:

Step 1: Set Up the Function Template​

Follow the instructions in the first step for setting up SMS trigger and name the function as WhatsAppTrigger. Set the authorization level to anonymous or function, depending on your security preference.

Step 2: Reference the Azure Communication Services Messages Package​

Ensure the Azure.Communication.Messages NuGet package is included in your project to enable messaging features needed for WhatsApp. Install the package with the following command in Visual Studio Code’s terminal:

bash

Add a reference to using Azure.Communication.Messages then create a property in the WhatsApp Trigger class to hold an instance of NotificationMessagesClient and a property to hold the WhatsApp identifier.

csharp

Step 3: Read Configuration and Initialize NotificationMessagesClient​

Update the WhatsAppTrigger class constructor to read the Azure Communication Services connection string from environment variables using Environment.GetEnvironmentVariable() and initialize NotificationMessagesClient with this connection string:

csharp

Step 4: Define the Request Model​

Create a request model class named WhatsAppRequest within the WhatsAppTrigger class, containing properties for the destination phone number, template name, language, and template parameters:

csharp

Step 5: Parse the Request Body​

Convert the Run function to be async to enable asynchronous work. Use StreamReader to read the request body and deserialize it to a WhatsAppRequest instance using System.Text.Json.JsonSerializer with JsonNamingPolicy.CamelCase.

csharp

Handle potential deserialization failure by returning BadRequestResult:

csharp

Step 6: Prepare Template Message and Send WhatsApp Message​

Modify the try-catch block to construct a SendMessageOptions object using MessageTemplateWhatsAppBindingsand MessageTemplate, and then make a call to _messagesClient.SendMessageAsync(sendTemplateMessageOptions):

csharp

Step 7: Return Success Response​

After sending the WhatsApp message successfully, return an OkObjectResult stating “WhatsApp sent successfully!”.

csharp

Final Code​

Following the described steps, the final WhatsAppTrigger Azure Function should look like this:

csharp

The WhatsAppTrigger Azure Function is now ready to send WhatsApp template messages. Be sure to test it extensively and remember to handle any issues related to input validation and communicate with the Azure Communication Services API correctly.

Deployment and Testing​

After developing the multichannel notification system using Azure Functions, the next step is to deploy and test the functions. This section will guide you through deploying your Azure Function to the cloud and testing the Email, SMS, and WhatsApp triggers.

Deploying the Azure Function​

Deployment of your Azure Function can be done right from Visual Studio Code with the Azure Functions extension.

Publish the Function App: In Visual Studio Code, sign in to Azure if you haven’t already. In the Azure Functions extension tab, find the ‘Deploy to Function App…’ button and select it.

Choose Your Function App: You can either create a new Function App or deploy it to an existing one. If it’s the first time you are deploying, choose ‘Create New Function App in Azure…’.
Set the Configuration: Provide a unique name for your Function App, select a runtime stack (.NET Core in this case), choose the appropriate region, and confirm your selections.
Wait for Deployment: The deployment process will take a few minutes. Monitor the output window for completion status and any potential errors.

Set Up Application Settings​

After deployment, you need to configure the application settings (environment variables) in Azure.

Open the Function App: Navigate to the Azure Portal, and find your Function App under ‘All Resources’ or by searching the name you provided.
Access Application Settings: In the Function App’s menu, go to ‘Configuration’ under the ‘Settings’ section.
Add the Settings: Click on ‘New application setting’ and add the key-value pairs for the environment variables specified in your local.settings.json: COMMUNICATION_SERVICES_CONNECTION_STRING, SENDER_EMAIL_ADDRESS, SENDER_PHONE_NUMBER, WHATSAPP_NUMBER, etc.,

json

Save and Restart: After adding the required settings, make sure to save the configurations and restart the Function App to ensure the new settings take effect.

Alternatively, when the Function has finished deploying, you can click on ‘Upload settings’ to upload your settings from local.settings.json. Don’t forget to restart the Function App after uploading the settings.

Testing the Function​

With the deployment complete and the environment configured, it’s time to verify that your function works as intended through each communication channel.

Testing Email Notifications​

To test the EmailTrigger function:

Send an HTTP POST Request: Use a tool like Postman to send a POST request to the Function App’s URL suffixed with /api/EmailTrigger. The body should contain JSON with keys for subject, htmlContent, and recipient.
Verify Email Receipt: Check the recipient’s email inbox for the message. Ensure that the subject and content match what you sent through the POST request.

Testing SMS Notifications​

To test the SMSTrigger function:

Send an HTTP POST Request: Using Postman, send a POST request to the Function App’s URL with /api/SMSTrigger at the end. The body of your request should contain JSON with message and phoneNumberkeys.
Check for SMS: Ensure that the specified phone number receives the SMS and the message content matches the request.

Testing WhatsApp Notifications​

To test the WhatsAppTrigger function:

Send an HTTP POST Request: Use Postman again to POST to the Function URL, this time ending with /api/WhatsAppTrigger. Include a JSON body with keys for phoneNumber, templateName, templateLanguage, and templateParameters.
Confirm WhatsApp Message: Verify that the WhatsApp message reaches the intended recipient with correct template filling.

Integrate with OpenAI GPTs​

In OpenAI GPTs editor, click ‘new GPT’ and ‘configure’. Name it “Email Sender” and set the description and instructions as mentioned.

Add Actions and JSON Schema​

Click ‘Create new action’ in your GPT configuration. Enter the following JSON:

json

Leave Authentication to none, and Privacy Policy blank.

Test Your GPT​

Finally, try out your GPT in the preview pane to see it in action!

By following these steps, you can easily integrate Azure Communication Services with OpenAI GPTs to send emails effortlessly.

Conclusion and Further Reading​

We have successfully walked through the journey of building a serverless multichannel notification system using Azure Functions and Azure Communication Services. This system can send timely and personalized notifications across multiple channels, such as Email, SMS, and WhatsApp. In addition, we have explored how to enhance our system with sophisticated content generation capabilities using OpenAI GPTs.

The modular nature of the Azure Functions framework allows your application to scale and adapt easily to changing requirements and traffic demands. Meanwhile, Azure Communication Services enrich the user experience by meeting customers on their preferred platforms, contributing to a seamless and cohesive communication strategy.

As developers, there’s always room to expand our knowledge and add robust features to our applications. Here are some suggestions for further exploration and resources that can assist you in taking your applications to the next level:

Azure Communication Services AI samples: One stop shop for GitHub samples for AI-powered communication solutions.
Azure Functions Best Practices: Learn about best practices for designing and implementing Azure Functions by visiting Azure Functions best practices.
Azure Communication Services Documentation: Explore the full capabilities of Azure Communication Services including chat, phone numbers, video calling, and more on the Azure Communication Services documentation.
Security and Compliance in Azure: Understand the best practices for security and compliance in Azure applications, particularly relevant for handling sensitive user communication data. Check the Microsoft Azure Trust Center.
OpenAI GPT Documentation: For more insight into using and customizing OpenAI GPTs, refer to the OpenAI API documentation.
Azure AI Services: Azure offers a range of AI services beyond just communication. Explore Azure AI services for more advanced scenarios such as speech recognition, machine translation, and anomaly detection at Azure AI services documentation.
Handling Large-scale Data: To handle a large amount of data and improve the performance of communication systems, consider learning about Azure’s data-handling services like Azure Cosmos DB, Azure SQL Database, and Azure Cache for Redis. Start with the Azure Data storage documentation.
Monitoring and Diagnostics: Improve the reliability of your applications by implementing robust monitoring and diagnostics tools. Azure offers several tools such as Azure Monitor and Application Insights. Dive into Application Insights for Azure Functions.
Serverless Workflow Automation with Azure Logic Apps: Enhance your serverless applications using Azure Logic Apps to automate and simplify workflows. Learn more about Azure Logic Apps at What is Azure Logic Apps?.

 Happy coding!

​Microsoft Tech Community – Latest Blogs –Read More 

This post is authored in conjunction with Yatish Anand, Senior Solutions Architect at Databricks, and Andrey Mirskiy, Senior Specialist Solutions Architect at Databricks.

Figure 1. Power BI automatic aggregations overview, source.

Introduction

In today’s fast-paced data landscape, timely insights can make all the difference. Power BI’s Automatic Aggregations feature is a breakthrough, designed to push performance boundaries by delivering low-latency query results on large datasets. It harnesses AI-powered caching to streamline DirectQuery models, combining the best-in-class performance of Power BI on Azure Databricks with low-latency BI for modern reporting needs. Used with DirectQuery mode, Automatic Aggregations do not face data volume limitations and allow you to scale regardless of data size without compromising on BI performance.

This innovation makes it easy for Power BI users of all skill levels to tap into advanced performance without worrying about backend strain or complex data modeling. Imagine your reports updating in real-time, even with billions of records in play. With this approach, you can deliver actionable insights faster than ever, freeing up time to focus on your business, not your data infrastructure. In this blog, we will showcase the integration of Power BI Automatic Aggregation with Azure Databricks and how this integration will help improve the performance of your Power BI reports.

What Are Automatic Aggregations?

Automatic aggregations streamline the process of improving BI query performance by maintaining an in-memory cache of aggregated data. This means that a substantial portion of report queries can be served directly from this in-memory cache instead of relying on the backend data sources. Power BI automatically builds these aggregations using AI based on your query patterns and then intelligently decides which queries can be served from the in-memory cache and which are routed to the data source through DirectQuery, resulting in faster visualizations and reduced load on the backend systems.

Key Benefits of Automatic Aggregations

Faster Report Visualizations: Automatic aggregations optimize most report queries by caching aggregated query results in advance, including those generated when users interact with reports.  Only outlier queries that cannot be resolved via the cache are directed to the data source.
Balanced Architecture: Compared to using pure DirectQuery mode, automatic aggregations enable a more balanced approach. Most frequently used queries are served from Power BI query in-memory cache, which reduces the processing load on data sources during peak reporting times, improves scalability, and decreases costs.
Simplified Setup: Model owners can easily activate automatic aggregations and schedule regular refreshes. Once the initial training and refresh are complete, the system autonomously develops an aggregation framework tailored to the specific queries and data patterns.

Configuring Automatic Aggregations

Setting up automatic aggregations is straightforward. Users can enable the feature in the model settings and schedule one or more refresh operations. It is essential to review comprehensive guidelines on how automatic aggregations function to ensure they are suitable for your specific environment.

Figure 2. Enabling automatic aggregations, source.

Once configured, Power BI will utilize a query log to track user interactions and optimize the aggregations cache over time. The training operation, which evaluates query patterns, occurs during the first scheduled refresh, allowing Power BI to adapt to changing usage patterns.

Requirements for Automatic Aggregations

Automatic aggregations are compatible with several Power BI plans, including:

Power BI Premium per capacity
Fabric F Sku Capacity
Power BI Premium per user
Power BI Embedded models

Automatic aggregations are specifically designed for DirectQuery models, including composite models that utilize both import tables and DirectQuery connections.

Automatic Aggregation Walkthrough with Azure Databricks Integration

In this example we will showcase how to enable Automatic Aggregations on Power BI semantic models and train Automatic Aggregations in order to enhance the performance of reports using Azure Databricks as a data source.

Pre-requisites

Before you begin, ensure you have the following:

An Azure Databricks account, access to an Azure Databricks workspace, and a Databricks SQL Warehouse.
Power BI Desktop installed on your machine. The latest version is highly recommended.
Power BI workspace
DAX Studio or any other DAX parser tool

Step by Step Instructions

1. Create an initial Power BI semantic model based on samples catalog, tpch schema. Add tables and relationships as shown on the screenshot below. The dimension tables customer and nation should be set to Dual storage mode. The fact tables orders and lineitem should be set to DirectQuery storage mode. Below is the data model for the sample report.

For Best practices around Power BI storage mode please refer to this Git repo

2. Create a simple tabular report showing the count of orders and min shipment date, sum of discounts and sum of quantities. Also add the slicer with nation names, as shown below.

3. Now publish this report to a Power BI workspace.

4. As shown below when we run the report it Power BI takes ~20 sec to run the query. Below is the snapshot from the Network Trace:

Also below screenshot shows query hit the Databricks SQL Warehouse and read 38M records.

5. Enable the Automatic Aggregations in the semantic model settings. You can set the Query coverage according to your needs. This setting will increase the number of user queries analyzed and considered for performance improvement. The higher percentage of Query coverage will lead to more queries being analyzed, hence higher potential benefits, however, aggregation training will take longer.

6. For Power BI to be able to create aggregations, we need to populate the Power BI query log which stores internal queries created by Power BI when users interact with a report. Thus, you can either open the deployed Power BI Report and interact with the report by selecting different nation names in the slicer or you can open the DAX studio and run the sample DAX query mentioned below.
Please note that for better model training you need to set different values for the slicer or the filter in DAX-query and run it multiple times.

TREATAS({“BRAZIL”}, ‘nation'[n_name])

One of the guidelines to populate a query log is that before making a report  available to users , the report publisher should open the report and try with different slicer filters. In our scenario as mentioned above we populated the query log by selecting the different names in the report slicer . This step would help end user have faster report rendering.

7. You can now start the model training manually or schedule it.

8. Once the model is trained, Power BI will have aggregated values in in-memory cache. The next time you interact with the report using similar patterns (dimensions, measures, filters) Power BI will leverage cached aggregations to serve the queries and will not send queries to Databricks SQL Warehouse. Hence, you may expect sub-second report refresh performance.

As shown in the below screenshot post enabling Automatic Aggregation we can see that the report visual is now getting rendered in ~1.6 sec as compared to 20 sec earlier. This is because the data is now getting read from query log cache.

Also as shown below there is no SQL query fired at the DBSQL as well

Monitoring and Managing Automatic Aggregations

Power BI continuously refines the in-memory aggregations cache through scheduled refreshes. Semantic model owners can choose to trigger training operations on demand if necessary. It’s also crucial to monitor the refresh history to ensure operations complete successfully and to identify any potential issues.

Power BI provides detailed refresh history logs that display the performance of each operation, enabling users to keep track of memory usage and other critical metrics

Conclusion

In today’s data-driven world, the integration of Azure Databricks and Power BI Automatic Aggregations is a game-changer, delivering unparalleled performance for even the most demanding data environments. While Azure Databricks excels at processing multi-terabyte datasets, Automatic Aggregations uses AI on your query patterns to intelligently cache aggregates, dramatically accelerating performance and reducing costs. This combination addresses the limitations of Import and Direct Lake modes, which are limited at working with large volumes while enhancing the efficiency of DirectQuery models. As shown in our blog with Automatic Aggregation on DirectQuery models you can now get sub-second report performance without constantly querying the underlying data source. With this innovative approach, you can focus on delivering lightning-fast BI reports at any scale rather than manually tuning your semantic model.

​Microsoft Tech Community – Latest Blogs –Read More 

Hello!

I am running a Linux VM on Azure (IaaS) which is providing an SFTP service to the Internet. Sadly, many customers are connecting to this service via public IP address (as opposed to FQDN).

I am migrating this service back to on-premises, through a firewall on a different public IP address.

Linux VM has public IP 1.1.1.1 right on its NIC.

Firewall’s IP is 2.2.2.2.

I want to redirect traffic to the on-premises firewall.

Is there an Azure service/resource that can take inbound connections to 1.1.1.1, then NAT the destination IP to 2.2.2.2 and then also NAT the source IP to 1.1.1.1 or another public IP (like 3.3.3.3) on that service/resource?

Thanks!

​Hello! I am running a Linux VM on Azure (IaaS) which is providing an SFTP service to the Internet. Sadly, many customers are connecting to this service via public IP address (as opposed to FQDN). I am migrating this service back to on-premises, through a firewall on a different public IP address. Linux VM has public IP 1.1.1.1 right on its NIC.Firewall’s IP is 2.2.2.2.   I want to redirect traffic to the on-premises firewall. Is there an Azure service/resource that can take inbound connections to 1.1.1.1, then NAT the destination IP to 2.2.2.2 and then also NAT the source IP to 1.1.1.1 or another public IP (like 3.3.3.3) on that service/resource? Thanks!  Read More

​