Azure Stack Reviews

My role has been primarily as an architect, utilizing Azure Stack for building data lakes, data warehouses, real-time solutions, and incorporating a little bit of AI for tasks like image recognition and voice-to-text. I am transitioning towards using AI and ML with technologies like ChatGPT.

The scalability of Azure Stack is commendable. It can scale up within two to three minutes upon request. SQL Server is a good use case for data warehouses. It is scalable and can be set to serverless to reduce costs when not in use. 

Power BI Fabric offers a useful way to stack data in data warehousing solutions. Azure Stack provides federated authentication which is seamless with on-premise infrastructure, offering good performance and throughput. The integration with Microsoft's environment is also efficient.

My primary use case for this solution is for cybersecurity and device management.

I find Cloud App Security and activity alerts particularly valuable. They allow me to track all user activities. Everything is stored under one roof, so there is no need for different licenses for different components.

We have our services platform built on Azure Stack. We are building our entire technology stack for serving business needs on Azure Stack. We are using Event Hubs, Event Grids, Synapse, and Azure App service.

It meets our requirements and we have continued with that stack. We did explore disaster recovery options, but currently we are satisfied with the solution. It is simpler to use this platform.

We use Azure Stack for repository management, storage, and services such as speech recognition, text-to-speech, and speech-to-text. These services are essential for the development lifecycle. You can check the relevant information on GitHub for application storage and CI/CD pipeline releases.

Face recognition for security in web applications can use AI services for processing. To store data, you can use SQL Azure, which involves exposing an API to access information from the face recognition service. You need to connect with Azure SQL Database and handle any issues related to JNC charts or time functions.

For example, you might photograph a person in front of a PC or laptop. This image can be processed using Azure face recognition services. The system compares the image against a repository in SQL Azure. Access can be granted to the application if there is a match with the stored data. If there's no match, the system will notify that the person is not an employee.

Azure Stack is a well-suited solution, especially for small to medium-sized businesses. While its scalability for large corporations is still untested, it is an excellent option for smaller enterprises looking to leverage the entire Azure platform. Azure offers a wide range of services, catering to various needs. For those interested in no-code or low-code automation, Microsoft provides the Power Platform, which includes tools like Power Apps for automation, Power Pivot, and more. These are ideal for teams or departments looking to streamline processes. If someone wants to use Azure as an enterprise data platform, they offer good solutions like Azure Data Lake, specifically using ADX Gen 2 for enterprise storage, which is commonly used. It pairs well with Databricks, especially when handling performance-intensive tasks, as Azure Synapse can be expensive for certain scenarios, such as data warehousing. For typical ETL tasks using Azure Data Factory and Azure Data Storage, the costs are quite reasonable.

I believe that Azure's Delta Lake features align with the Azure Lake House architecture, similar to the traditional Medallion architecture, which comprises bronze, silver, and gold layers. This structure is valuable for effective data warehouse management. You can store your raw data in the bronze layer, standardized data in the silver layer, and customized data in the gold layer. Tools like Informatica and Data Factory, which serve as ETL (Extract, Transform, Load) solutions, do a commendable job of connecting to diverse data sources, processing them, and loading them into Azure Data Lake Storage Gen2 (ADLS Gen2). Data Factory, in particular, seamlessly integrates with ADLS Gen2, making it a powerful combination for data processing. It offers the option to leverage Azure Databricks on top of this setup. Databricks provides the capability to work with data tables and allows for custom program development using Python, Spark, or SQL. Combining Azure's solutions with Databricks can lead to massive scalability and efficient processing of big data tasks.

Our businesses have diverse scopes, leading to a wide range of requirements that we use this solution for. These encompass everything from the initial onboarding process in an on-premises environment to aspects like security, analytics, and integration.

The two most crucial aspects of it revolve around connectivity and security. In terms of connectivity, it's vital to ensure seamless external connections, bridging on-premises and cloud environments through features such as site-to-site VPNs and Access Routes. On the other hand, we're prioritizing security, especially with the use of Azure Firewall, including its premium features, as well as SQM.

Our company provides local cloud services with multiple platforms and Azure Stack is one of our platforms.

The most valuable thing about Azure Stack Hub is that it provides the same control panel for managing cloud services and provides the same experience we have with global Azure, for the supported services.

We have used the solution for IPS, IDS, and security incident response. We have used Defender as an antivirus, anti-malware, and intrusion detection. We used Intune for endpoint management and endpoint detection response. We use Purview for insider threats, data leakage detection and response, and data classification.

You can use the Azure security policies to enforce your compliance objectives. We also used web firewalls and identity governance, which is a security feature or can be used as such.

The solution's front door and web application firewalls have proven most beneficial for our edge computing needs. The solution provides end-to-end comprehensive protection across the services.