Biometric Face Recognition is the process and ability of a bio metric machine to identify and recognize the face of an individual. It deals with either to grant access to a secured system or to find out the details of a person by matching the face with existing data in the machine’s system.
Facial recognition is full of potential and can be easily incorporated to increase the security measures of any device/object. Apart from all the excitement, this technology is still developing, the more faces are fed in the algorithm, the more accurate it becomes. Therefore, there’s no need to be afraid of facial technology as it is being used for good ethical uses and safe practices.
This post focuses on one of the solution with Azure by using various services such as Cognitive Service, Blob Storage, Event Grid, Azure Functions and Cosmosdb to build the right architecture that would solve the above mentioned use case.
The above architecture is very self explanatory, it comprises of two main components/flow.
Training Faces of Individuals
Identifying faces of Individuals
Every process is achieved in the above case using Azure function. Each operation will have separate function to achieve result. Main operations such as RegisterUser,TrainFace,TriggerTrain are simple Azure functions in the above diagram. Images are uploaded to Blob Storage using SAS token and face detection is done using cognitive service and references are stored in CosmosDB. EventGrid is used to route the events according to the invokes, for ex, to train image whenever a user uploads while registration and tag the face of the individual in the database.
I have used the above architecture with one case study and hope it will help someone out there who wants to build similar solution. Cheers!
Biometric Face Recognition is the process and ability of a bio metric machine to identify and recognize the face of an individual either to grant access to a secured system or to find out the details of a person by matching the face with existing data in the machine’s system.
Facial recognition is full of potential and can be easily incorporated to increase the security measures of any device/object. Apart from all the excitement, this technology is still developing, the more faces are fed in the algorithm, the more accurate it becomes. Therefore, there’s no need to be afraid of facial technology as it is being used for good ethical uses and safe practices.
Industries around the globe have already started to use face detection for several purposes. As a series of my ideas, continuation with Architecture for traffic problem, This post focus on one of the solution with Azure by using various services such as Cognitive Service, Blob Storage, Event Grid, Azure Functions and Cosmosdb to build the right architecture that would solve the above mentioned use case.
The above architecture is very self explanatory, it comprises of two main components/flow.
Training Faces of Individuals
Identifying faces of Individuals
Every process is achieved in the above case using Azure function. Each operation will have separate function to achieve result. Main operations such as RegisterUser,TrainFace,TriggerTrain are simple Azure functions in the above diagram. Images are uploaded to Blob Storage using SAS token and face detection is done using cognitive service and references are stored in CosmosDB. EventGrid is used to route the events according to the invokes, for ex, to train image whenever a user uploads while registration and tag the face of the individual in the database.
I have used the above architecture with one case study and hope it will help someone out there who wants to build similar solution. Cheers!
It's been almost a week since Angular 9 was released which included several new features such as new Ivy Renderer, SSR, Type checking etc. In this post, i will be going through on how to build and deploy an Angular9 application as a static website to Azure with GitHub Actions.
We will use the package by Wassim Chegham to deploy the application to Azure. Next step in your Angular project make sure to install the Hexa.run CLI as a prod dependency as you can see from the package.json of this project.
Next step is to create the necessary resources to deploy the angular application. In this way, we will deploy our application to static website for Azure Storage which is the optimum option to host a single page application (SPA) on Azure.Hosting a SPA in pure Storage is by far the cheapest and most efficient way of running in Azure.
You can login to azure account with the command,
npm run hexa:login
which will list down the available subscriptions and you need to pick the subscription where you want to deploy the application.
Next step is to initiate the configuration needed for the deployment of the application. Run the Hexa CLI command as follows,
npm run hexa:init
which will ask for few inputs from the user such as the project name, storage account name and the destination folder. Eventually, you will see a new file generated as hexa.json which will look like the below,
You need to use the service principal identity mechanism to do the authorization of the deployment. In order to generate the service principal using hexa, run the below command,
npm run hexa:ci
Hexa.run will automatically:
create an Azure resource group (or lets you choose an existing one)
create the Azure storage account
configure the storage account and make it static-website ready
upload the angular bundle.
prints the generated URL from the Azure service.
Also it will generate the necessary credentials as a JSON and make a note of them.
Now it's the time to create the Actions workflow, you can create a new workflow by navigating to actions and click New Worfklow. There are few sample template workflows available , in this case we will use own workflow. So you need to click on setup workflow yourself.
Setting up own workflow GitHub Actions
Immediately you could see new workflow.yml file created where you need to add the steps and actions needed to deploy the app. Here is the workflow file look like after adding all the steps.
name: Deploy to Azure with Hexa.ru on: push: branches: - master - release/* jobs: build: runs-on: ubuntu-latest strategy: matrix: node-version: [12.x] steps: - uses: actions/checkout@v1 - name: Use Node.js ${{ matrix.node-version }} uses: actions/setup-node@v1 with: node-version: ${{ matrix.node-version }} - name: npm install run: | npm install - name: npm build, and deploy env: AZURE_SERVICE_PRINCIPAL_ID: ${{ secrets.AZURE_SERVICE_PRINCIPAL_ID }} AZURE_SERVICE_PRINCIPAL_PASSWORD: ${{ secrets.AZURE_SERVICE_PRINCIPAL_PASSWORD }} AZURE_SERVICE_PRINCIPAL_TENANT: ${{ secrets.AZURE_SERVICE_PRINCIPAL_TENANT }} run: | npm run hexa:login npm run build -- --prod npm run hexa:deploy
As you could see, the steps are very much simple starting with installing the dependencies and deploying the angular application using the hexa:deploy command.
Also you need to configure the secrets in the github repository which were generated in the step 6. You can create a new secret by navigating to settings and then secrets. you need to define the below secrets which are associated with the service principal.
Github Secrets
The rest in the workflow can be easily understood as its about the environment and the trigger(whenever someone push the changes to master/release there should be a build)
Step 9 : See Github Actions in Action
Immediately when you save the workflow.yml you can see there will be a new build triggered and the steps are executed, which you can notice in the Actions tab as follows,
That's all you need to do inorder to deploy the Angular application to Azure. If you need to include end to end testing and different tasks you could simply modify the flow and it. Github Actions definitely a future to believe in! Try this out and let me know if you have any queries! Cheers!
It's been almost a week since Angular 9 was released which included several new features such as new Ivy Renderer, SSR, Type checking etc. In this post, i will be going through on how to build and deploy an Angular9 application as a static website to Azure with GitHub Actions.
We will use the package by Wassim Chegham to deploy the application to Azure. Next step in your Angular project make sure to install the Hexa.run CLI as a prod dependency as you can see from the package.json of this project.
Next step is to create the necessary resources to deploy the angular application. In this way, we will deploy our application to static website for Azure Storage which is the optimum option to host a single page application (SPA) on Azure.Hosting a SPA in pure Storage is by far the cheapest and most efficient way of running in Azure.
You can login to azure account with the command,
npm run hexa:login
which will list down the available subscriptions and you need to pick the subscription where you want to deploy the application.
Next step is to initiate the configuration needed for the deployment of the application. Run the Hexa CLI command as follows,
npm run hexa:init
which will ask for few inputs from the user such as the project name, storage account name and the destination folder. Eventually, you will see a new file generated as hexa.json which will look like the below,
You need to use the service principal identity mechanism to do the authorization of the deployment. In order to generate the service principal using hexa, run the below command,
npm run hexa:ci
Hexa.run will automatically:
create an Azure resource group (or lets you choose an existing one)
create the Azure storage account
configure the storage account and make it static-website ready
upload the angular bundle.
prints the generated URL from the Azure service.
Also it will generate the necessary credentials as a JSON and make a note of them.
Now it's the time to create the Actions workflow, you can create a new workflow by navigating to actions and click New Worfklow. There are few sample template workflows available , in this case we will use own workflow. So you need to click on setup workflow yourself.
Setting up own workflow GitHub Actions
Immediately you could see new workflow.yml file created where you need to add the steps and actions needed to deploy the app. Here is the workflow file look like after adding all the steps.
name: Deploy to Azure with Hexa.ru on: push: branches: - master - release/* jobs: build: runs-on: ubuntu-latest strategy: matrix: node-version: [12.x] steps: - uses: actions/checkout@v1 - name: Use Node.js ${{ matrix.node-version }} uses: actions/setup-node@v1 with: node-version: ${{ matrix.node-version }} - name: npm install run: | npm install - name: npm build, and deploy env: AZURE_SERVICE_PRINCIPAL_ID: ${{ secrets.AZURE_SERVICE_PRINCIPAL_ID }} AZURE_SERVICE_PRINCIPAL_PASSWORD: ${{ secrets.AZURE_SERVICE_PRINCIPAL_PASSWORD }} AZURE_SERVICE_PRINCIPAL_TENANT: ${{ secrets.AZURE_SERVICE_PRINCIPAL_TENANT }} run: | npm run hexa:login npm run build -- --prod npm run hexa:deploy
As you could see, the steps are very much simple starting with installing the dependencies and deploying the angular application using the hexa:deploy command.
Also you need to configure the secrets in the github repository which were generated in the step 6. You can create a new secret by navigating to settings and then secrets. you need to define the below secrets which are associated with the service principal.
Github Secrets
The rest in the workflow can be easily understood as its about the environment and the trigger(whenever someone push the changes to master/release there should be a build)
Step 9 : See Github Actions in Action
Immediately when you save the workflow.yml you can see there will be a new build triggered and the steps are executed, which you can notice in the Actions tab as follows,
That's all you need to do inorder to deploy the Angular application to Azure. If you need to include end to end testing and different tasks you could simply modify the flow and it. Github Actions definitely a future to believe in! Try this out and let me know if you have any queries! Cheers!
Microsoft's Power Platform is a low code platform which enable the organization to analyse data from multiple sources, act on it through application and automate business process. Power Platform contains 3 main aspects (Power Apps,Power BI and Microsoft Flow) and it also integrates two main ecosystems(Office 365 and Dynamics 365). In this blog we will see in detail about Power Apps which helps to quickly build apps that connects data and run on web and mobile devices.
We wills see how to build a simple Power App to automate scaling of resources in Azure in a simple yet powerful fashion to help save you money in the cloud using multiple Microsoft tools. In this post, we will use Azure Automation run books to code Azure Power Shell scripts, MS Flow as orchestration tool and MS Power Apps as the simple interface. If you have an Azure Environment with lot of resources it becomes hassle to manage the scaling part of it if you don't have the auto scaling implemented. The following Power App will help to understand how easy it is to build the above scenario.
First we need to create the Azure Automation account. Navigate to Create Resources -> search for Automation Account on the search bar. Create a new resource as follows,
Once you have created the resource, you also need to install the required modules for this particular demo which includes -AzureRM.Resources
-AzureRM.Profile
Next step is to create the run books by going to the Runbooks blade, for this tutorial lets create 6 run books one for each resources and its purpose. We need to create PowerShell scripts for each of those types
Scale Up SQL
-Scale Up ASP
-Start VM
-Scale Down SQL
-Scale Down ASP
-Stop VM
Creating RunBook
PowerShell scripts for each of these are found in the Github repository. We need to create runbook for each of the scripts.
All the runbooks above can be scheduled and automated to run for desired length of time, particular days of the week and time frame or continuously with no expiry. For an enterprise for non-production you would want it to scale down end of business hours and at the weekend.
Step 3: Create and Link Microsoft Flow with Azure Automation Run books
Once we tested with the above Powershell scripts and the runbooks, tested and published now we can move on the next step to create the Flow. Navigate to Flow and Create New App from the template.
New Flow from the template
Select the template as PowerApps Button and the first step we need to add is the automation job. When you search for automation you will see the list of jobs available. Select Create Job and pick the one you created above. If you want to all the actions in one app, you can add one by one, If not you need to create separate flows for each one.
In this one, i have created one with the ScaleUpDB job which will execute the Scale up command of the database.
Once you are done with all the steps save the flow with necessary name.
Step 4 : Create Power Apps and and link to your Flow button
Once we create PowerApp flow buttons login to MS PowerApps with a work/school account. Navigate to Power Apps which will give a way to create a blank canvas for Mobile or tablet. Next you can then begin to customise the PowerApp with text labels, colour and buttons as below
PowerApps Name
In this case we will have a button to increase/decrease the count of instances of the sql db, my app looked like below with few labels and buttons.
AutoMateApp
Next is to link the flow to the button of the power App by navigating to the Actions -> Power Automate
Linking Button Action
Once both Scale Up/Scale down actions are linked, save the app and publish
Inorder to verify if things are working correctly, click on scale up and scale down few times and navigate to Azure Portal and open the Automation account we created.
Navigate to the overview tab to see the requests for each job made via the power app as below.
Jobs executed.
In order to look at the history navigate to the Jobs blade
Jobs Execution
further you can build a customized app for different environments with different screens using Power App. With the help of Azure Alert, whenever you get an alert regarding the heavy usages of resources/spikes, with single click of button you will be able scale up and scale down the resources as you need.
This is just a starting point to explore more on this functionality, but there are improvements you could add to make this really useful.
(i) Sometimes the azure automation action fails to start the runbook. When you are implementing flow needs to handle this condition.
(ii) Sometimes a runbook action will be successful, but the runbook execution errored. Consider using try/catch blocks in the PowerShell and output the final result as a JSON object that can be parsed and further handled by the flow.
(iii) We should update your code to use the Az modules rather than AzureRM.
Note : The user who executes the PowerApp also needs permission to execute runbooks in the Automation Account.
With this App, It becomes handy for the operations team to manage the portal without logging in. Hope it helps someone out there! Cheers.
