Achieving sustainable growth with Azure and FinOps best practices

Companies worldwide are committed to reducing their IT carbon footprint, championing a more sustainable future through initiatives focused on efficiency and cost optimization. Cloud sustainability is not only about reducing the environmental impact of cloud usage, but also about making smart business decisions that align to corporate values, adhere to regulatory requirements, and enable the pursuit of long-term business goals. To understand the impact of cloud computing on carbon emissions, precise measurement, trustworthy data, and robust tools are essential.

That’s why we’re excited to announce two new capabilities to optimize your Microsoft Azure emissions:  

• Azure Carbon Optimization (preview) is a free, cutting-edge capability that empowers Azure developers and IT professionals to understand and optimize emissions stemming from Azure usage. By providing insights into carbon emissions and offering recommendations for enhancing cloud efficiency, this tool aligns with the Microsoft commitment to environmental responsibility and supports you in achieving your cloud sustainability goals. 
• Microsoft Azure emissions insights (preview) in sustainability data solutions in Microsoft Fabric enables you to unify and analyze emissions data for Azure usage. By having access to your Azure emissions data in Microsoft Fabric, you can query and drill down into Azure resource level emissions for advanced reporting and analysis.

Both tools offer a holistic solution for organizations aiming to reduce their carbon footprint by optimizing specific resources or workloads within Azure.  

With Azure Carbon Optimization (preview), engineering and IT teams can use ready-to-consume insights and recommendations for optimizing their carbon emissions, all within the Azure portal. Microsoft Azure emissions insights (preview) enable data analysts and engineers to dive deeper into emissions data, allowing them to slice and dice the data and perform deeper analytics using Microsoft Fabric. 

Once your organization can access insights into the carbon emissions generated at the resource or workload level, reduction efforts can begin. This involves optimizing cloud systems for efficiency to benefit the environment and enhance overall performance. Azure administrators can already see a company-wide view of cloud emissions in the Emissions Impact Dashboard. To optimize your carbon footprint, you can take advantage of more granular insights into carbon emissions originating from specific resources or workloads. 

Like any major organizational shift, reducing carbon emissions requires contributions from every corner of your company. In this blog, we will not only explore the benefits of Azure Carbon Optimization and Microsoft Azure emissions insights, but also how the FinOps framework can guide your business through the complexities of carbon emission reduction to help achieve both your environmental and financial goals.  

Align IT sustainability with ESG regulations

Organizations around the world are setting carbon neutrality goals for themselves, which are furthered by new environmental regulations and standards introduced by global government and regulatory bodies, with a significant driver being environmental, social, and governance (ESG) regulations. These governmental standards dictate ESG-related actions, reporting, and disclosures.

Microsoft provides offerings to help customers with their ESG reporting needs with tools and products available with Microsoft Cloud for Sustainability to help your organization collect and manage more ESG data and get fuller visibility into your environmental impact. Our goal is to help prepare you for any new reporting requirements by compiling a comprehensive ESG data estate. 

IT sustainability plays a pivotal role in a company’s ESG management strategy because it serves as a cornerstone for mitigating environmental impact, ensuring responsible cloud usage, and reinforcing the overall commitment to sustainable development practices. There are also direct economic benefits for reducing carbon emissions, such as long-term operational cost savings.  

Above all, organizations that proactively address environmental issues and reduce their carbon footprint will be better positioned for long-term success, especially in a business landscape where sustainability is increasingly important.  

Measure and reduce your emissions with Azure Carbon Optimization

Our free Azure Carbon Optimization tool, now in public preview and accessible through your Azure portal, is a window into your cloud resources emissions, ultimately leading to recommendations on how to cut back. It empowers Azure users to closely monitor and optimize their carbon footprint. 

Azure Carbon Optimization is designed to provide everyone in your organization, from developers, to architects, to IT professionals, with a resource-level view of emissions data. This empowers your engineers to take proactive measures to mitigate emissions and track progress right from the Azure portal.  

Azure Carbon Optimization uses the same carbon accounting methodology as the Emissions Impact Dashboard. Developers can work towards maximizing resource utilization while minimizing carbon emissions from the cloud, helping ensure that every deployed resource serves a purpose, eliminates waste, and reduces environmental impact. The tool also presents carbon emission reduction in equivalent terms that are easy for anyone to understand. 

Subsequently, it provides developers with carbon savings recommendations that are based on analyzing resource utilization. Suggestions include deleting or resizing underutilized resources. With these ready-to-consume recommendations, you can optimize your Azure usage, avoid carbon emissions, and promote sustainable development practices. This way, you not only enhance your environmental performance, but also achieve cost savings and efficiency.

Perform even deeper Azure emissions analysis with Microsoft Fabric

Microsoft Azure emissions insights, now in public preview, is a part of the sustainability data solutions in Microsoft Fabric. It helps unify, process, query, and perform deeper analysis of Azure emissions data. In addition to emissions data and related pipelines, Power BI dashboards are provided with Microsoft Azure emissions insights to drill-down and compare emissions data across subscriptions and resources. This helps IT administrators identify patterns in Azure emissions that evolve with time and change with Azure resource usage. 

Unified Azure emissions data empowers data analysts to enrich the emissions data with custom information such as department using subscriptions and resources. They can then query the data and build analytic models for interesting insights such as Azure emissions by departments and seasonality of emissions by usage.  

Leverage FinOps best practices to help optimize carbon emissions

Fostering a culture of accountability, efficiency, and governance across an organization stands as a key objective within the FinOps framework, which aims to help organizations optimize their cloud to maximize business value.  

Efficiency has a positive impact on innovation by freeing up resources and allowing organizations to invest more in modernization, research, and development. FinOps supports the customer journey by establishing a cross-functional team that includes finance, IT, engineers, and business leaders to create a culture of accountability where everyone takes ownership of their cloud usage.  

As ESG regulations compel adherence to complex emissions reporting requirements, integrating FinOps best practices can help teams to better manage and optimize carbon emissions. When viewed through the lens of environmental awareness, FinOps can assist with best practices that foster accountability, efficiency, and governance to enable data-driven decisions.  

Leveraging these best practices in tandem with Azure Carbon Optimization and Microsoft Azure emissions insights empowers your organization to be a catalyst for change, transforming cloud practices into a force for sustainability by helping track, analyze, and optimize emissions towards a greener, more responsible cloud ecosystem.  

Reach your sustainability goals with data-driven Azure insights

By employing these capabilities and adhering to FinOps practices, your organization can actively track, assess, and mitigate your carbon emissions. You’ll not only gain a detailed understanding of the emissions impact associated with your Azure resources, but also valuable insight into your compliance posture for any coming ESG regulations.

Source: microsoft.com

Continue reading

Dynatrace and the Microsoft commercial marketplace: AI-powered cloud transformation

In today’s digital landscape, the cloud has become a cornerstone for over 90% of organizations. This widespread adoption brings a growing complexity to cloud portfolios as companies navigate the challenges of migrating on-premises environments, managing hybrid workloads, and modernizing their cloud estate—all while operating under current business imperatives of speed and scale. To overcome this, organizations need to arm themselves with advanced tools that streamline manual processes as well as offer insightful analytics.

Microsoft Azure is the world’s most trusted cloud, allowing customers to build intelligent apps and solutions while seamlessly unifying technology to simplify platform management and deliver innovations efficiently and securely. To support customers in getting the most from their cloud portfolio, Microsoft works with partners such as Dynatrace. Dynatrace, an analytics and automation platform powered by causal AI, was recognized as a Leader in the 2023 Gartner® Magic Quadrant™ for Application Performance Monitoring and Observability.

Azure and Dynatrace, better together

With Dynatrace, customers can simplify their IT ecosystems at scale to adjust to market conditions more quickly. The Azure Native Dynatrace Service simplifies the process for businesses to migrate and optimize their cloud estate, and because the solution is available through the Microsoft commercial marketplace, customers can click-to-deploy and adjust investments based on usage.

Dynatrace recently unveiled its latest advancement in cloud management: the Dynatrace Grail Data Lakehouse. This innovative technology is part of the Azure Native Dynatrace Service, designed to adeptly manage vast amounts and types of data across cloud-native, hybrid, and multicloud environments. Grail unifies observability, security, and business data while maintaining crucial context, delivering precise and efficient analytics instantly and cost-effectively.

One of Grail’s standout features is its capability to process various logs and events without the traditional burdens of indexing, re-indexing, or schema management. This removes the preliminary requirements of creating custom scripts, tags, and indexes before data ingestion, streamlining the process significantly.

The Dynatrace platform is fully compatible with all Azure services, including Azure Spring Cloud, Azure Kubernetes Service, Azure Cosmos DB, Azure Linux, and Azure Functions. This compatibility enhances its functionality and provides near real-time monitoring capabilities.

AI-centric innovation with Dynatrace

AI is at the core of the Dynatrace platform, using predictive, causal, and generative AI together to provide root-cause analysis and deliver precise answers, intelligent automation, accurate predictions, and increased productivity. The ability to leverage AI and AIOps ensures customers can accelerate innovation and maximize automation as IT environments continue to evolve.

“Thriving in the digital era requires a unified approach to observability, security, and business data analytics. Bringing our latest platform technologies to Microsoft Azure enables more customers and teams within organizations to harness our industry-leading AI, analytics, and automation capabilities to modernize cloud operations, expedite releases of high-quality and secure software, and ensure flawless digital experiences for their users.”

—Steve Tack, SVP of Product Management, Dynatrace.

Reducing time-to-value through the Microsoft commercial marketplace

Dynatrace offerings are readily accessible through the Microsoft commercial marketplace, streamlining the process from acquisition to implementation. Marketplace simplifies the journey for businesses seeking the right cloud solutions, offering a user-friendly platform for discovery, purchase, trial, and deployment. It’s a one-stop shop where customers can explore and select from a variety of vetted solutions, like Dynatrace, that seamlessly integrate with existing technology stacks.

As IT budgets grow with increased reliance on the cloud, customers are seeking better value from their investments. Microsoft helps its customers achieve more by automatically counting eligible marketplace purchases towards their Azure consumption commitment—a contract that empowers customers to get discounts on their cloud infrastructure after agreed upon spending thresholds are achieved.

“Many of our customers have cloud commits, and we’ve identified the [Microsoft] commercial marketplace as a significant opportunity for them to become more efficient with their IT budgets and increase growth by using those commits to purchase Dynatrace.” 

—Ayla Anderson, Microsoft Alliance Manager, Dynatrace.

Many Microsoft customers are turning to the marketplace as their go-to resource for managing their cloud portfolios efficiently and confidently while empowering customers to buy how they want. The marketplace supports many buying scenarios, whether it’s a digital purchase, a private offer from a solution provider, like Dynatrace, or a deal procured by their channel partner with a multiparty private offer.

Recently, IT strategy firm Trace3 procured Dynatrace’s solution on their customer’s behalf with a multiparty private offer sold through the marketplace. Because Dynatrace’s solution is Azure benefit eligible, the purchase counted towards the customer’s Azure consumption commitment, and they received the benefits of simplified purchasing while getting more value from their cloud dollars.

Source: microsoft.com

Continue reading

The Total Economic Impact™ of the Microsoft commercial marketplace

In today’s fast-paced, AI-fueled business environment, organizations require a growing number of cloud solutions. In trying to meet this demand, software as a service (SaaS) sprawl can lead to wasted investments and cost overruns. Research by Forrester found that cloud marketplaces can minimize this challenge by helping organizations balance speed and agility. The Microsoft commercial marketplace allows customers to centralize their cloud portfolio, taking complexity out of the purchasing process. In a 2023 Total Economic Impact™ study commissioned by Microsoft, Forrester Consulting found the Microsoft commercial marketplace delivered customers a 587% return on investment (ROI) with a payback period of less than six months.

By connecting customers to our ecosystem of more than 400,000 partners through one trusted source, the marketplace simplifies and accelerates business-to-business (B2B) tech commerce. Forrester Consulting interviewed 10 customers that actively use the Microsoft commercial marketplace to quantify the opportunity. The research found that for a composite organization based on the interviewed customers the marketplace:

• Reduced employee time for onboarding new vendors by 75%.
• Reduced employee effort required for each procurement engagement by 50%.
• Optimized Azure cloud budget by recapturing up to 50% of at-risk spend.

“I can actually do procurement faster. I can make sure that my products get released faster, bringing benefits faster to the company.”

Director of IT, Financial Services

Simplify procurement processes

Enterprise buying is complex, especially in tech and with rapid advancements in AI. Procurement teams must navigate a multitude of choices, manage vendor relationships, and continuously adapt to emerging trends while ensuring investments meet immediate needs, as well as align with long-term business strategy in a cost-efficient manner.

Ultimately, organizations find the most success with the marketplace when the procurement team, or buying office, becomes involved at an organizational level. As a single destination to discover, try, and purchase certified solutions, the marketplace helps customers “make informed procurement decisions without having to devote significant employee time to researching, evaluating, and selecting a vendor,” according to interviewees in the Forrester Consulting study. Using the marketplace reduced employee effort required for each procurement engagement by 50%.

“We have readily available options for our suppliers that are already vetted and available. [Our teams] can adopt without having to go through elongated procurement processes, additional orders, and administrative processes.”

Director of Global Supply Chain, Telecommunications

For finance teams, interviewees reported in the Forrester Consulting study that leveraging features of the marketplace such as automated invoicing and billing systems “reduced manual data entry, eliminated billing errors, and expedited overall invoice processing and billing cycles.” Adopting the marketplace improved time savings for the payments and finance purchasing team by 30%.

“Once I buy a solution from the marketplace, I don’t have the nightmare of managing invoices and checking for accuracy.”

Executive Director, Government

Onboarding new vendors can be slow and cumbersome, at times delaying a customer from deploying new technology solutions. Using the marketplace reduced time required to onboard new vendors by 75%. Vendors can submit required documentation and access guidance and tools directly in the marketplace, reducing the administrative burden of reviewing paperwork and walking vendors through new processes.

“The Microsoft commercial marketplace has been really useful for fulfilling my day-to-day activities and facilitating my vendor relationships. …The contract and red tape is handled by Microsoft. I don’t have to go deep into my NDAs (non-disclosure agreements), legal documents, or contract redlining, and all those kinds of things.”

Director of IT, Financial Services

Maximize cloud investments

As cloud investments grow, customers are looking to maximize value. Many of the customers interviewed by Forrester Consulting for the study had consumption commitments with Microsoft Azure—contracts providing discounts on infrastructure once a customer meets a certain spending threshold.

The Microsoft commercial marketplace is unique in that it automatically counts 100% of eligible purchases towards a customer’s commitment, ensuring more value for the investment. Customers can quickly search Azure benefit-eligible solutions directly in the Azure portal.

Because of this benefit, Forrester Consulting found the marketplace helped customers optimize their Azure commitments, recapturing up to 50% of at-risk spend. Those interviewed by Forrester Consulting are in good company. According to Microsoft data, more than 85% of customers with consumption commitments are already buying partner solutions via the marketplace to maximize their investments.

“It’s dollar for dollar. Whatever we spend, 100% of it goes to our spend commitment, and that was a huge deal. For some of [Microsoft’s] competitors, [that’s] not the case.”

IT Sourcing Specialist, Manufacturing

Marketplace customers also got more out of their cloud budgets by avoiding costly proof of concept fees. Prior to using the marketplace, customers interviewed by Forrester Consulting said they often had to pay up-front fees for a proof of concept to vet new solutions. The marketplace makes it simple to access trial options, including free trials ranging from 30 days to six months, so customers can feel confident before they buy.

Source: microsoft.com

Continue reading

Reflecting on 2023—Azure Storage

The beginning of a new year often prompts reflection along with planning for the future. At the forefront of our priorities lies the commitment to enhance the Azure platform and its ecosystem, fuel groundbreaking AI solutions, and facilitate seamless migrations to the cloud. We achieve this through purpose-built storage solutions tailored to meet the unique demands of your workloads.

2023 was an emblematic year of growth for storage and data services with AI at the top of mind for many customers. Azure customers’ overall data estate continued to grow, powered by the emergence of new workloads and access patterns. Notable Azure Storage figures include:

• The storage platform now processes more than 1 quadrillion (that’s 1000 trillion!) transactions a month with over 100 exabytes of data read and written each month. Both numbers were sharply higher compared to the beginning of 2023.
• Premium SSD v2 disks, our new general-purpose block storage tailored for SQL, NoSQL databases, and SAP grew capacity by 100 times.
• The total transactions of Premium Files, our file offering for Azure Virtual Desktop (AVD) and SAP, grew by more than 100% year over year (YoY).

In 2023, advancements and investments were made in five strategic areas aligning with customer workload patterns, cloud computing trends, and the evolution of AI.

Focused innovations for new workloads

We’ve made advancements in offering end-to-end solutions with unique storage capabilities, tailored to help customers bring new workloads to Azure, without the need to retrofit them.

As customers modernize applications with Kubernetes, we have observed customers’ workloads moving from “stateless” to “stateful” workloads. Azure Container Storage is custom designed to meet the needs of hosting stateful workloads where storage is tightly integrated with containers and offers outstanding data management, lifecycle management, price-performance, and scale. Azure Container Storage is the industry’s first platform-managed container-native storage service in the public cloud. It offers a path to manage persistent volumes and the backend storage options completely via the Kubernetes native experience.

Microsoft’s Azure Elastic SAN is a fully managed storage area network (SAN) service that simplifies deploying, scaling, managing, and configuring a SAN in the cloud. It is designed to address customers’ challenges in service management when it comes to migrating a large-scale SAN appliance to cloud. We are the first in the market in providing a fully managed SAN offering on cloud, with built-in cloud scale and high availability options. It adopts a SAN-like hierarchy, provisioning resources at the top-level resource, dynamically sharing resources across the workloads, managing security policies at the workload level. In addition, it enables customers to bring new data intensive workloads to Azure VMware Solution using Elastic SAN to serve the storage needs. Elastic SAN and Azure Container Storage, in public preview, showcase a paradigm shift by simplifying service management, purpose built for cloud native storage scenarios. General availability is planned for the upcoming months.

AI is leading a fresh wave of innovation powered by data. Azure Blob Storage stands at the forefront of this data explosion, providing excellent performance and high bandwidth access for all AI training needs. Azure Premium Block Blob, generally available since 2019, provides extremely low latency at competitive pricing for AI and machine learning (ML), IoT and streaming analytics, and interactive workloads. We also released Azure Managed Lustre in July which offers a fully managed distributed parallel file system tailored for high performance computing (HPC) and AI training workloads.

OneLake (which is the storage layer for Microsoft Fabric) is powered by the scale, performance, and rich capabilities of Azure Data Lake Storage (ADLS). ADLS has been a game changer for analytics with differentiated storage features including native file system capabilities that enable atomic meta data operations, resulting in a significant performance boost, especially when used with Azure Premium Block Blob. Multi-protocol access is another unique capability that allows customers to work with their data using a variety of access protocols (including REST, SFTP and NFS v3) eliminating data silos, costly data movement, and duplication.

Optimizations for mission critical workloads

We continue to add differentiated capabilities to enable enterprises to optimize their mission critical workloads hosted on Azure.  

With the general availability of Premium SSD v2, we have been actively onboarding solutions focused on mission critical data. Gigaom highlighted the superior performance and reduced total cost of ownership of SQL Server 2019 running on Azure Virtual Machines (VMs) with Premium SSD v2. Similar benefits apply to other database workloads like PostgreSQL. We will continue enhancing Premium SSD v2 capabilities and expand regional coverage in the upcoming year.

Azure Blob Storage powers data-intensive mission critical workloads globally, including some of the largest AI model training and inferencing. Cloud-native companies host exabytes of consumer-generated data and rely on the platform’s scale-out capabilities. Major retailers store trillions of user activities to power recommendation engines and inventory systems. The Cold storage tier, released in 2023, enables cost-efficient retention of this data.

Enterprises migrate Windows Server workloads to the cloud for modernization, supported by enhanced Azure Files and Azure File Sync capabilities. The preview of Microsoft Entra ID (previously named as Azure Active Directory support for Azure Files REST API) enables modern applications to use identity-based authentication and authorization for accessing SMB file shares. Azure Files offers comprehensive authentication support, including on-premises AD and Microsoft Entra ID for both Kerberos and REST. Unique to Azure, we also offer Microsoft Entra ID Kerberos authentication enabling seamless migration from on-premises Windows servers to Azure which keeps your data and permissions intact while eliminating complex domain join setups in the cloud. Enterprises migrating mission-critical applications such as SAP and Oracle rely on Azure NetApp Files (ANF) to meet their performance and data protection requirements.  In strategic partnership with NetApp, we introduced the public preview of ANF large volumes up to 500 TiB, addressing the needs of workloads that require larger capacity under a single namespace, such as HPC in Electronic Design Automation and Oil and Gas related applications, and the public preview of Cool Access, improving the TCO for infrequently accessed data.

Efficient data migration at scale is critical for successful cloud workload migration. We offer a comprehensive suite of migration solutions. Azure Storage Mover, a fully managed data migration service, improves the overall migration experience of large data sets. Azure File Sync allows hybrid synchronization between on-premises Windows Servers and Azure Files. AzCopy, a command-line utility, enables quick script-based migration jobs, while Azure Databox facilitates offline data transfer.

Expanding partner ecosystem

We transition customer workloads to Azure in collaboration with our partners, who complement our platform capabilities with innovative solutions. We have fostered numerous successful strategic partnerships and are delighted to see the partner community investing in exclusive Azure solutions.

Qumulo announced the next version of Azure Native Qumulo Scalable File Service, exclusive to Azure and built to leverage Azure Blob Storage’s scale and cost efficiency. This service supports exabytes of data under a single namespace, surpassing other public cloud file solutions, and facilitates high-scale enterprise network attached storage (NAS) data migration to Azure.

Commvault unveiled a purpose-built cyber resilience solution, exclusively on Azure. It empowers users to predict threats faster, achieve clean recoveries, and accelerate response times. It integrates seamlessly with Microsoft Azure OpenAI Service.

Since the launch of Pure Storage Cloud Block Storage offering on Azure in 2021, we have partnered closely with Pure Storage. They use unique Azure Storage features such as shared disks and were early adopters of Ultra and Premium SSD v2 disk storage. Our collaboration has expanded to containers through Portworx and the Azure VMware Solution.

Besides storage and backup partners, we collaborate with Atempo, Data Dynamics, Komprise, and Cirrus Data to deliver storage migration solutions, providing access to industry-leading file and block solutions to migrate petabytes of NAS and storage area network (SAN) data to Azure at no additional charge.

Industry contributions

We actively participate in industry organizations such as Storage Networking Industry Association (SNIA) and the Open Compute Platform (OCP), collaborating with industry leaders to share insights, influence standards, and contribute to the development of innovative solutions and best practices.

At the 2023 Storage Developer Conference, we presented Massively Scalable Storage for Stateful Containers on Azure, emphasizing the synergy between Azure Container Storage and Elastic SAN. This combination offers unified volume management across diverse storage backends and delivers a highly available block storage solution designed to scale to millions of input/output operations per second (IOPS) with fast pod attach and detach experience.

At Flash Memory Summit (FMS) conference we chartered the course for the future of flash as a HDD displacement technology in alignment with Azure Storage long-range plans.

We are committed to sustainability, striving for net carbon-zero goals in the next decade. Through the OCP Sustainability Project, we provide an open framework for the datacenter industry to adopt best practices for reusability and circularity and lead in the steering committee to drive progress.

Unparalleled commitment to quality

Our primary focus is on delivering a robust foundation for our storage services, adapting to evolving workloads like the emergence of AI and ML. This ongoing commitment covers reliability, durability, scalability, and performance, for our storage solutions while maintaining a low total cost of ownership (TCO) for Azure customers.

We assure customers of high data durability, backed by substantial investments in infrastructure, hardware, software, and streamlined processes. Our Zone-redundant Storage (ZRS) ensures data durability of at least 99.9999999999% (12 9’s) over a given year—the highest standard among the major cloud service providers (CSPs). Notably, we are the sole cloud provider among the major CSPs maintaining a 0% Annual Failure Rate (AFR) with our block storage offerings since launch.

Our unique ZRS offering provides a simple way to establish highly available solutions across three zones at a fraction of the costs. You can deploy a cross-zonal AKS cluster with persistent volumes hosted on ZRS disks to ensure data durability and availability during zonal outages. Clustered applications like SQL failover cluster instances (FCI) leveraging Windows Server Failover Cluster (WSFC) can also benefit from ZRS with high resiliency provided out of the box.

In a concerted effort to elevate customer support experience, we’ve integrated with Azure Copilot, empowering support engineers to troubleshoot more efficiently while delivering quality responses quickly. This has resulted in a noteworthy increase in customer satisfaction scores.

In closing, 2023 has been a year of profound learning and substantial progress. Explore our feature releases in Azure Updates. Our commitment is to empower you with Azure Storage innovations and enable the seamless execution of a diverse range of workloads. We invite you to continue trusting us as we anticipate exciting developments in 2024!

Source: microsoft.com

Continue reading

Elevate Azure expertise with new AI and optimization video episodes

The Azure Enablement Show is a library of fun, bite-sized videos that highlight valuable resources for anyone looking to learn more about cloud computing and related Azure tools. In this blog we announce two new video series: Azure Optimization Skilling, which explores best practices and resources to maximize the value of your cloud investment, and Develop Your AI Coding Skills, which discusses how AI-enabled applications have become crucial to meet the evolving demands of users.

Organizations of all sizes are embracing cloud computing to enhance their operations, optimize resource utilization, and gain a competitive edge. However, navigating the complexities of cloud adoption and ensuring optimal performance can be a daunting task. This is where the Azure Enablement Show steps in, providing valuable resources for businesses seeking to unlock the full potential of cloud computing.

These short videos, featuring industry experts from Microsoft and the broader Azure community, give a snapshot of new resources and programs to help guide you in your Azure journey. In this blog, we’ll introduce two new Azure Enablement Shows to further enhance your Azure expertise.

Learn to optimize and secure your Azure investment with new skilling programs

The benefits of cloud computing—improved scalability, growth prospects, innovation potential, and more—are widely acknowledged as valuable assets for businesses. The path to confident cloud operations includes cost efficiency, reliability, security, and sustainability. Azure Optimization aims to address this by fostering the ongoing advancement of our customers’ cloud investment.

To maximize the value of your Azure cloud investment through optimization, you need to reduce unnecessary expenses and enhance your ROI. We understand the importance of improving workload reliability, security, and cloud cost optimization for accelerating business innovation without concerns about vulnerabilities or escalating costs. Our goal is to support the continual improvement of your Microsoft Azure workloads, allowing you to consistently extract maximum value from your cloud investment once key optimization elements are in place.

Whether you’re a seasoned cloud architect, an aspiring developer, or a business decision-maker eager to explore the possibilities of cloud computing, optimization should be at the forefront of your cloud strategy. We help prepare you for your cloud optimization journey with training resources and guidance across all stages and levels of maturity.

In the Azure Enablement Show on Optimization Skilling, you’ll watch a quick video overview about our curated learning opportunities, such as our instructor-led Virtual Training Days, a 30 Days to Learn It Optimization Challenge, and self-guided online learning paths and modules. You’ll also discover how optimizing workloads and architecture can lead to growth and innovation and hear from experts about the best practices and tools they recommend from years of experience.

We aim to boost your confidence and efficiency in achieving your cloud goals by providing diverse learning paths and skilling opportunities, offering resources such as the Microsoft Cloud Adoption Framework (CAF) and the Well-Architected Framework (WAF). These tools empower you to transform operations, securely manage cloud-hosted applications, and achieve business goals cost-effectively.

Raise the bar in app development with AI coding skills

App development is a highly competitive field, and today’s most dynamic apps combine personalization with a user-friendly experience that makes them more engaging. Intelligent apps integrate AI and machine learning capabilities to understand user preferences and behavior, tailoring content, recommendations, and user interfaces accordingly. Such personalization enhances the overall user experience, making apps more engaging. 

AI has transformed how we work and interact with technology, and developers with the skills to enable AI in their apps will have a distinct advantage in the marketplace. With our three-part Azure Enablement Show series, Develop Your AI Coding Skills, you’ll discover how adding AI skills to your developer toolbox will help both you and your organization.

Enabling AI in app development offers benefits that significantly enhance the landscape of modern applications. By allowing developers to create intelligent, responsive applications that stand out in a crowded market, AI-driven features not only attract users but also elevate the overall user experience.

The integration of AI also enables performance optimization, ensuring applications operate seamlessly, adapt to user behavior, and continuously improve functionality. This results in faster, more efficient applications that meet the evolving demands of users. In addition, AI opens new and exciting opportunities for innovation. Developers can explore novel features, personalized experiences, and predictive capabilities that go beyond conventional app functionalities, thereby unlocking untapped potential and reshaping the possibilities within the realm of app development.

Microsoft Azure delivers on the potential of intelligent apps with an innovative suite of AI and machine learning tools, designed to streamline development and enhance user experiences. In this video series we’ll explore how developers can take advantage of our comprehensive learning resources on Microsoft Learn to harness the full capabilities of Azure and drive innovation in their applications.

Check out the full lineup of Azure Enablement Shows

Azure Enablement Shows are a treasure trove of knowledge and insights designed to empower businesses of all sizes to maximize their cloud investments and achieve their strategic objectives. By tuning in to these informative sessions, you’ll gain the knowledge and guidance you need to navigate the complexities of cloud adoption, build robust and scalable cloud applications, and optimize your cloud solutions for peak performance and cost-effectiveness. Join a community of cloud enthusiasts, engage with experts from Microsoft and the broader Azure community, and immerse yourself in a world of technical guidance, tips, and best practices. Azure Enablement Shows are not just about learning; they’re about unlocking the full potential of cloud computing to transform your business. 

Source: microsoft.com

Continue reading

Achieve generative AI operational excellence with the LLMOps maturity model

In our LLMOps blog series, we’ve explored various dimensions of Large Language Models (LLMs) and their responsible use in AI operations. Elevating our discussion, we now introduce the LLMOps maturity model, a vital compass for business leaders. This model is not just a roadmap from foundational LLM utilization to mastery in deployment and operational management; it’s a strategic guide that underscores why understanding and implementing this model is essential for navigating the ever-evolving landscape of AI. Take, for instance, Siemens’ use of Microsoft Azure AI Studio and prompt flow to streamline LLM workflows to help support their industry leading product lifecycle management (PLM) solution Teamcenter and connect people who find problems with those who can fix them. This real-world application exemplifies how the LLMOps maturity model facilitates the transition from theoretical AI potential to practical, impactful deployment in a complex industry setting.

Exploring application maturity and operational maturity in Azure

The LLMOps maturity model presents a multifaceted framework that effectively captures two critical aspects of working with LLMs: the sophistication in application development and the maturity of operational processes.  

Application maturity: This dimension centers on the advancement of LLM techniques within an application. In the initial stages, the emphasis is placed on exploring the broad LLM capabilities, often progressing towards more intricate techniques like fine-tuning and Retrieval Augmented Generation (RAG) to meet specific needs.  

Operational maturity: Regardless of the complexity of LLM techniques employed, operational maturity is essential for scaling applications. This includes systematic deployment, robust monitoring, and maintenance strategies. The focus here is on ensuring that the LLM applications are reliable, scalable, and maintainable, irrespective of their level of sophistication. 

This maturity model is designed to reflect the dynamic and ever-evolving landscape of LLM technology, which requires a balance between flexibility and a methodical approach. This balance is crucial in navigating the continuous advancements and exploratory nature of the field. The model outlines various levels, each with its own rationale and strategy for progression, providing a clear roadmap for organizations to enhance their LLM capabilities. 

LLMOps maturity model 

Level One—Initial: The foundation of exploration 

At this foundational stage, organizations embark on a journey of discovery and foundational understanding. The focus is predominantly on exploring the capabilities of pre-built LLMs, such as those offered by Microsoft Azure OpenAI Service APIs or Models as a Service (MaaS) through inference APIs. This phase typically involves basic coding skills for interacting with these APIs, gaining insights into their functionalities, and experimenting with simple prompts. Characterized by manual processes and isolated experiments, this level doesn’t yet prioritize comprehensive evaluations, monitoring, or advanced deployment strategies. Instead, the primary objective is to understand the potential and limitations of LLMs through hands-on experimentation, which is crucial in understanding how these models can be applied to real-world scenarios. 

At companies like Contoso1, developers are encouraged to experiment with a variety of models, including GPT-4 from Azure OpenAI Service and LLama 2 from Meta AI. Accessing these models through the      catalog allows them to determine which models are most effective for their specific datasets. This stage is pivotal in setting the groundwork for more advanced applications and operational strategies in the LLMOps journey. 

Level Two—Defined: Systematizing LLM app development 

As organizations become more proficient with LLMs, they start adopting a systematic method in their operations. This level introduces structured development practices, focusing on prompt design and the effective use of different types of prompts, such as those found in the meta prompt templates in Azure AI Studio. At this level, developers start to understand the impact of different prompts on the outputs of LLMs and the importance of responsible AI in generated content.

An important tool that comes into play here is Azure AI prompt flow. It helps streamline the entire development cycle of AI applications powered by LLMs, providing a comprehensive solution that simplifies the process of prototyping, experimenting, iterating, and deploying AI applications. At this point, developers start focusing on responsibly evaluating and monitoring their LLM flows. Prompt flow offers a comprehensive evaluation experience, allowing developers to assess applications on various metrics, including accuracy and responsible AI metrics like groundedness. Additionally, LLMs are integrated with RAG techniques to pull information from organizational data, allowing for tailored LLM solutions that maintain data relevance and optimize costs.  

For instance, at Contoso, AI developers are now utilizing Azure AI Search to create indexes in vector databases. These indexes are then incorporated into prompts to provide more contextual, grounded and relevant responses using RAG with prompt flow. This stage represents a shift from basic exploration to a more focused experimentation, aimed at understanding the practical use of LLMs in solving specific challenges.

Level Three—Managed: Advanced LLM workflows and proactive monitoring  

During this stage, the focus shifts to refined prompt engineering, where developers work on creating more complex prompts and integrating them effectively into applications. This involves a deeper understanding of how different prompts influence LLM behavior and outputs, leading to more tailored and effective AI solutions.  

At this level, developers harness prompt flow’s enhanced features, such as plugins and function callings, for creating sophisticated flows involving multiple LLMs. They can also manage various versions of prompts, code, configurations, and environments via code repositories, with the capability to track changes and rollback to previous versions. The iterative evaluation capabilities of prompt flow become essential for refining LLM flows, by conducting batch runs, employing evaluation metrics like relevance, groundedness, and similarity. This allows them to construct and compare various metaprompt variations, determining which ones yield higher quality outputs that align with their business objectives and responsible AI guidelines. 

In addition, this stage introduces a more systematic approach to flow deployment. Organizations start implementing automated deployment pipelines, incorporating practices such as continuous integration/continuous deployment (CI/CD). This automation enhances the efficiency and reliability of deploying LLM applications, marking a move towards more mature operational practices.  

Monitoring and maintenance also evolve during this stage. Developers actively track various metrics to ensure robust and responsible operations. These include quality metrics like groundedness and similarity, as well as operational metrics such as latency, error rate, and token consumption, alongside content safety measures.  

At this stage in Contoso, developers concentrate on creating diverse prompt variations in Azure AI prompt flow, refining them for enhanced accuracy and relevance. They utilize advanced metrics like Question and Answering (QnA) Groundedness and QnA Relevance during batch runs to constantly assess the quality of their LLM flows. After assessing these flows, they use the prompt flow SDK and CLI for packaging and automating deployment, integrating seamlessly with CI/CD processes. Additionally, Contoso improves its use of Azure AI Search, employing more sophisticated RAG techniques to develop more complex and efficient indexes in their vector databases. This results in LLM applications that are not only quicker in response and more contextually informed, but also more cost-effective, reducing operational expenses while enhancing performance. 

Level Four—Optimized: Operational excellence and continuous improvement  

At the pinnacle of the LLMOps maturity model, organizations reach a stage where operational excellence and continuous improvement are paramount. This phase features highly sophisticated deployment processes, underscored by relentless monitoring and iterative enhancement. Advanced monitoring solutions offer deep insights into LLM applications, fostering a dynamic strategy for continuous model and process improvement. 

At this advanced stage, Contoso’s developers engage in complex prompt engineering and model optimization. Utilizing Azure AI’s comprehensive toolkit, they build reliable and highly efficient LLM applications. They fine-tune models like GPT-4, Llama 2, and Falcon for specific requirements and set up intricate RAG patterns, enhancing query understanding and retrieval, thus making LLM outputs more logical and relevant. They continuously perform large-scale evaluations with sophisticated metrics assessing quality, cost, and latency, ensuring thorough evaluation of LLM applications. Developers can even use an LLM-powered simulator to generate synthetic data, such as conversational datasets, to evaluate and improve the accuracy and groundedness. These evaluations, conducted at various stages, embed a culture of continuous enhancement.  

For monitoring and maintenance, Contoso adopts comprehensive strategies incorporating predictive analytics, detailed query and response logging, and tracing. These strategies are aimed at improving prompts, RAG implementations, and fine-tuning. They implement A/B testing for updates and automated alerts to identify potential drifts, biases, and quality issues, aligning their LLM applications with current industry standards and ethical norms. 

The deployment process at this stage is streamlined and efficient. Contoso manages the entire lifecycle of LLMOps applications, encompassing versioning and auto-approval processes based on predefined criteria. They consistently apply advanced CI/CD practices with robust rollback capabilities, ensuring seamless updates to their LLM applications. 

At this phase, Contoso stands as a model of LLMOps maturity, showcasing not only operational excellence but also a steadfast dedication to continuous innovation and enhancement in the LLM domain.

Identify where you are in the journey 

Each level of the LLMOps maturity model represents a strategic step in the journey toward production-level LLM applications. The progression from basic understanding to sophisticated integration and optimization encapsulates the dynamic nature of the field. It acknowledges the need for continuous learning and adaptation, ensuring that organizations can harness the transformative power of LLMs effectively and sustainably.

The LLMOps maturity model offers a structured pathway for organizations to navigate the complexities of implementing and scaling LLM applications. By understanding the distinction between application sophistication and operational maturity, organizations can make more informed decisions about how to progress through the levels of the model. The introduction of Azure AI Studio that encapsulated prompt flow, model catalog, and the Azure AI Search integration into this framework underscores the importance of both cutting-edge technology and robust operational strategies in achieving success with LLMs. 

Source: microsoft.com

Continue reading

Unwrapping the 2023 holiday season: A deep dive into Azure’s DDoS attack landscape

As the holiday season of 2023 unfolded, it brought not only cheer and celebration but also a surge in Distributed Denial-of-Service (DDoS) attacks. This year’s trends in DDoS attacks reveal a complex and evolving threat landscape. From misconfigured Docker API endpoints enabling botnet delivery to the emergence of NKAbuse malware exploiting blockchain technology for DDoS attacks, the tactics and scale of these attacks have shown significant sophistication and diversification.

The 2023 holiday season attack landscape in Azure

In our monitoring of the attack landscape during the holiday season, we observed a notable shift in some of the attack patterns compared to the previous year. This change underscores the relentless efforts of malicious actors to refine their threat tactics and attempt to circumvent DDoS protection strategies.

Daily Attack Volume: Azure’s robust security infrastructure automatically mitigated a peak of 3,500 attacks daily. Notably, large-scale attacks, exceeding 1 million packets per second (pps), constituted 15%-20% of these incidents.*

Figure 1: Number of daily DDoS attacks towards resources in Azure.

Geographical origins: A shift in attack origins was observed, with the top two origin countries being China with 42% of the attacks and the USA with 18%. All other countries make up 40% of attacks.* This marks a change from the previous year, where both countries were equally represented as the top two regional sources.

Figure 2: Source countries for DDoS Attacks on Azure.

Attack protocols: The 2023 holiday season saw a predominant use of UDP-based attacks, targeting gaming workloads and web applications, accounting for 78% of the attacks. These include UDP reflected/amplified attacks, which predominantly leverage domain name system (DNS) and simple service discovery protocol (SSDP), as well as quick UDP internet connections (QUIC) for reflection purposes. Notably, QUIC is emerging as a more common attack vector, either by reflection or by DDoS stressors that utilize UDP port 443 randomly. This year’s holiday season attack patterns contrast sharply with the previous year, where TCP-based attacks dominated 65% of all attacks.*

Figure 3: Attacks protocols distribution.

Record-breaking attack: A staggering UDP attack, peaking at 1.5 terabits per second (Tbps), targeted a gaming customer in Asia. This attack, originating from China, Japan, the USA, and Brazil, was highly randomized, involving numerous source IPs and ports, yet was fully mitigated by Azure’s defenses.

Botnet evolution: In the past year, cybercriminals increasingly leveraged cloud resources, particularly virtual machines, for DDoS attacks. This trend continued to evolve during the holiday season, with attackers trying to exploit discounted Azure subscriptions globally. From mid-November 2023 and until end of year, we monitored compromised account attempts in 39 Azure regions, with Europe and the USA being the primary targets, accounting for about 67% of these incidents.* Azure’s defense mechanisms successfully neutralized these threats.

Figure 4: Azure regions where attempts to exploit resources for DDOS attacks occurred.

Contextualizing the threat

The 2023 DDoS attack trends in Azure mirror global patterns. Attacks are becoming politically motivated as we highlighted earlier last year, fueled by geopolitical tensions.

The emergence of DDoS-for-hire services, commonly known as “stressers” and “booters” remain popular amongst attackers. These platforms, readily available on cybercriminal forums, have democratized the ability to launch powerful DDoS attacks, making them accessible to less sophisticated criminals for minimal costs. Recent years have seen an uptick in the availability and use of these services, confirmed by international law enforcement agencies through operations like Operation PowerOFF, which last year in May targeted 13 domains associated with DDoS-for-hire platforms. Despite these efforts, stressers continue to thrive, offering a range of attack methods and power, with some capable of attacks up to 1.5 Tbps.

Cloud power: Combating the evolving DDoS threats

The rise of botnets at scale and DDoS-for-hire services poses a significant risk to online services and business operations. To fight these threats, more cloud computing power is needed to absorb the leading wave of the attack until patterns can be identified, spurious traffic diverted, and legitimate traffic preserved. When tens of thousands of devices constitute an attack, the cloud is our best defense, due to the scale needed to mitigate the largest attacks. In addition, due to the global distribution of the cloud, closer proximity helps to block attacks closest to the sources.

Ensuring robust protection

In an era where digital threats are constantly evolving, ensuring robust protection against DDoS attacks has never been more critical. Here’s how Azure’s comprehensive security solutions are designed to safeguard your digital infrastructure.

DDoS Protection Service: With the high risk of DDoS attacks, it’s essential to have a DDoS protection service like Azure DDoS Protection. This service provides always-on traffic monitoring, automatic attack mitigation upon detection, adaptive real-time tuning, and full visibility on DDoS attacks with real-time telemetry, monitoring, and alerts.

Multi-Layered Defense: For comprehensive protection, set up a multi-layered defense by deploying Azure DDoS Protection with Azure Web Application Firewall (WAF). Azure DDoS Protection secures the network layer (Layer 3 and 4), while Azure WAF safeguards the application layer (Layer 7). This combination provides protection against various types of DDoS attacks.

Alert Configuration: Azure DDoS Protection can identify and mitigate attacks without user intervention. Configuring alerts for active mitigations can keep you informed about the status of protected public IP resources.

2024: Rising against DDoS threats

The 2023 holiday season has underscored the relentless and evolving threat of DDoS attacks in the cyber landscape. As we transition into the new year, it becomes crucial for organizations to enhance and adapt their cybersecurity strategies. This period should be a learning curve, focusing on fortifying defenses against such DDoS attacks and staying vigilant against new tactics. The resilience of Azure against these sophisticated DDoS threats highlights the critical need for robust and adaptive security measures, not just in protecting digital assets but also in ensuring uninterrupted business operations.

* Based on internal data

Source: microsoft.com

Continue reading