News For Dev-ops

Artificial intelligence (AI) has the potential to transform nearly every sector of the economy and society. For U.S. government agencies, the latest advances in AI offer a unique opportunity to optimize their operations, deliver better user experiences and enhance national security. Microsoft is innovating in AI to deliver technology for the public sector. As we embark on this new way to work, we are committed to delivering transformational AI innovations responsibly to drive government missions, aligned to the goals of the White House’s recent executive order for safe, secure and trustworthy AI. We began enabling government agencies to adopt new AI technologies for mission critical solutions with Azure OpenAI Service in the Azure commercial cloud. Azure OpenAI Service is included as a service within the FedRAMP High authorization for our commercial cloud, enabling many agencies to securely access new large language models (LLMs) like GPT 3.5 and GPT 4 for their unclassified (non-CUI...

With the recent arrival of the  Public preview of Workload identity federation for Azure Pipelines , you may be wondering how to efficiently migrate dozens or even hundreds of ARM Service Connections to take advantage of these benefits.  In this post, Emmanuel Knafo dives right in. Updating Your Azure DevOps ARM Service Connections To Use The Recommended Workload Identity Federation With the recent arrival of the  Public preview of Workload identity federation for Azure Pipelines , you may be wondering how to efficiently migrate my dozens or even hundreds of ARM Service Connections to take advantage of these main benefits: Simplified management : You do not need to generate, copy, and store secrets from service principals in Azure Entra ID to Azure DevOps anymore. Secrets that are used in other authentication schemes of Azure service connections (e.g., service principal) expire after a certain period (2 years currently). When they expire, pipelines fail. You have to g...

John Folberth  and  Joe Fitzgerald  share sample guidance for developing and deploying an Azure Data Factory into multiple environments. Introduction   Part 1 Architecture and Scenario Creating resources in Azure Create Azure Storage Containers Create Azure Key Vaults Create Azure Data Factory: With Key Vault Access Part 2  Configure Azure Data Factory Source Control Construct Azure Data Factory Data Pipeline Publishing Concept for Azure Data Factory Configure Deployed Azure Resources. Part 3 The YAML Pipeline Structure The Publish Process ARM Template Parameterization ADF ARM Template Deployment Part 4 How to use Azure DevOps Pipeline Templates This section will build on that and include how to build/deploy larger Data Factories which leverage  Azure Resource Manager Linked Templates  for deployment.     When Linked Templates Are Required   Typically, users will not desire to use linked templates; however, ...

PowerShell 7.4 RC-1 is now available We're proud to announce the availability of PowerShell 7.4.0-rc.1 ! This is the first release candidate version of PowerShell 7.4 and is considered a "go-live" release meaning that it is a supported release in production. Now is the time to test out PowerShell 7.4 in your environment. PowerShell 7.4 is built on top of .NET 8 and as a LTS (Long Term Support) release will be supported for 3 years. How do I get it? Since PowerShell 7 is supported on Windows, Linux, and macOS, there are a variety of ways to get it . If you had installed the previous PowerShell 7.4 preview release (7.4.0-preview.6) via the Windows Store or MSI (and opted into Microsoft Update ), you will be automatically updated to 7.4.0-rc.1. Note that after releases there are some delays in all release channels getting the latest bits. If this version is not available on your release channel of choice expect it to be available within 2 weeks. What's new in this rel...

Amazon Kinesis Data Streams now supports On-Demand write throughput limit of 2 GB/s, a 2x increase from the current limit of 1 GB/s. Amazon Kinesis Data Streams is a serverless streaming data service that makes it easier to capture, process, and store streaming data at any scale. On-Demand is a capacity mode for Kinesis Data Streams that automates capacity management, so that you never have to provision and manage the scaling of resources. It requires you to pay for throughput consumed rather than for provisioned resources, making it easier to balance costs and performance while providing the same availability, durability, and integrations. You can create a new On-Demand data stream or convert an existing data stream into the On-Demand mode with a single-click without requiring any code changes or downtime for your existing applications. Post Updated on October 30, 2023 at 05:19PM

As the Microsoft 365 service evolves, we review opportunities to improve the experiences of both users and developers. Aligned to the vision, we are moving to a unified search solution that spans all of Microsoft 365 and announcing that we will no longer be making improvements to the OneDrive and SharePoint search APIs in favor of Microsoft Search APIs in Microsoft Graph . This decision will allow focused development of the best search experience possible in a single coherent API. This change should encourage you to plan code updates, though we are not announcing an end-of-life date for the existing APIs. However, we will no longer invest in improvements or updates to these existing APIs. Our guidance is to transition your applications to call the Microsoft Graph Search API , and you will find examples of how to find OneDrive and SharePoint content later in this article. The existing OneDrive and SharePoint search APIs we will no longer be improving include: graph.microsoft.com/{ver...

A customer noted that the documentation for the Set­Prop function says, Before a window is destroyed (that is, before it returns from processing the WM_ NC­DESTROY message), an application must remove all entries it has added to the property list. The application must use the Remove­Prop function to remove the entries. What are the dire consequences of failing to remove properties? If you forgot to remove the property, the system will call Remove­Prop on your behalf, but it will also generate a debug message to remind you of your oversight. Removing properties is good for hygiene. It prevents the system from worrying that maybe you forgot something. Often, the value associated with the property is something that itself needs to be cleaned up, so if you fail to clean it up yourself, that's an indication that you leaked something. The text about cleaning up properties has been around for a very long time. I can't prove it, but it's possible that early versions of Wi...

Hello prompt engineers, I’ve been reading about how to improve the process of reasoning over long documents by optimizing the chunking process (how to break up the text into pieces) and then summarizing before creating embeddings to achieve better responses. In this blog post we’ll try to apply that philosophy to the Jetchat demo’s conference chat, hopefully achieving better chat responses and maybe saving a few cents as well. Basic RAG embedding When we first wrote about building a Retrieval Augmented Generation (RAG) chat feature, we created a ‘chunk’ of information for each conference session. This text contains all the information we have about the session, and it was used to: Create an embedding vector that we compare against user queries, AND Add to the chat prompt as grounding context when there is a high vector similarity between the embeddings for the chunk and the user query. Figure 1 shows an example of how the text was formatted (with key:value pairs) and the type...

Today, AWS announces multi-session capability for Amazon AppStream 2.0 fleets powered by Microsoft Windows server operating system. This new feature enables IT admins to host multiple end-user sessions on a single AppStream 2.0 instance, helping to make better use of instance resources. Post Updated on October 26, 2023 at 10:11PM

Starting today, Amazon EC2 I4i instances are now available in two additional instance sizes i4i.12xlarge and i4i.24xlarge in 28 AWS regions: US East (N. Virginia, Ohio), US West (N. California, Oregon), Canada (Central), Europe (Ireland, Frankfurt, Stockholm, Paris, London, Milan, Zurich), Asia Pacific (Hong Kong, Mumbai, Jakarta, Hyderabad, Melbourne, Osaka, Seoul, Singapore, Sydney, Tokyo), Middle East (Bahrain, UAE), Africa (Cape Town), South America (Sao Paulo), and AWS GovCloud (US-East, US-West).  Post Updated on October 26, 2023 at 10:00PM

AWS Network Firewall now supports egress Transport Layer Security (TLS) inspection, enabling customers to strengthen their security posture on AWS by improving visibility into encrypted outbound VPC traffic. Starting today, you can use AWS Network Firewall to decrypt, inspect, and re-encrypt outbound TLS traffic destined for the internet, another VPC, or another subnet. Post Updated on October 26, 2023 at 10:54PM

Amazon OpenSearch Service now offers customers the option to use Internet Protocol version 6 (IPv6) addresses for their new and existing domains. Customers moving to IPv6 can simplify their network stack by running their OpenSearch Service domains on a network that supports both IPv4 and IPv6. Post Updated on October 26, 2023 at 10:47PM

Problem Statement Application deployments in distributed environments with many clusters have always been troublesome. This includes cases like deployments on edge machines spread over geographical regions or even deployments of various components of distributed applications across co-located but substantial number of clusters. Kubernetes as a tool has come up as de-facto choice for orchestrating these deployments in easy and readable YAML (Yet Another Markup Language)L based deployment descriptors and GitOps tools like Flux, ArgoCD, and Jenkins X have emerged as popular choices for automating the deployment and management of these Kubernetes application, while these patterns work well for small number of clusters but the organization, configuration and scalability poses significant challenges with increasing number of clusters. A subset of these challenges is as below Determining application State and deployment locations: As the ultimate source of truth, the gitOps manifests become...

Starting today, Amazon Elastic Compute Cloud (Amazon EC2) M2 Mac instances are now generally available (GA). These instances deliver up to 10% faster performance over the existing M1 Mac instances when building and testing applications for Apple platforms such as iOS, macOS, iPadOS, tvOS, watchOS, visionOS, and Safari. M2 Mac instances are powered by the AWS Nitro System and are built on Apple M2 Mac Mini computers featuring 8 core CPU, 10 core GPU, 24 GiB of memory, and 16 core Apple Neural Engine. Post Updated on October 25, 2023 at 10:48PM

AWS Config now supports 19 more resource types for services, including Amazon AppStream 2.0, AWS Batch, AWS CodeBuild, Amazon CodeGuru Profiler, AWS Cloud Map, AWS Elastic Container Service (Amazon ECS), AWS Elemental MediaConnect, Amazon Inspector, AWS IoT, AWS IoT TwinMaker, AWS IoT Wireless, AWS Managed Service for Prometheus, Amazon Managed Streaming for Apache Kafka (Amazon MSK), Amazon Personalize, Amazon Route 53 Resolver, Amazon SageMaker, and AWS Transfer Family. Post Updated on October 25, 2023 at 10:07PM

Following the announcement of u pdates to the PostgreSQL database by the open source community, we have updated Amazon Aurora PostgreSQL-Compatible Edition to support PostgreSQL 15.4, 14.9, 13.12, 12.16, and 11.21. These releases contains product improvements and bug fixes made by the PostgreSQL community, along with Aurora-specific improvements. This release also contains new features and improvements such as Babelfish for Aurora PostgreSQL version 3.3 . Refer to the Aurora version policy to help you to decide how often to upgrade and how to plan your upgrade process. As a reminder, if you are running any version of Amazon Aurora PostgreSQL 11,  you must upgrade to a newer major version by February 29, 2024.  Post Updated on October 24, 2023 at 11:00PM

Today, AWS CodePipeline announces the general availability of parameterized pipelines, a new feature that enables customers to dynamically pass input parameters to a pipeline execution. For a pipeline to accept input parameters, customers have to declare the variables and use them in action configurations. When releasing a change through the console, customers can set values for variables to use in the pipeline execution. These input variables can be used in configuration of all action types except Source action types. Customers can also set the values for the input parameters when starting a pipeline through API, CLI, SDK and CDK. Post Updated on October 24, 2023 at 11:02PM

Amazon Aurora PostgreSQL-Compatible Edition now supports v0.5.0 of the pgvector extension to store embeddings from machine learning (ML) models in your database and to perform efficient similarity searches. This version includes Hierarchical Navigable Small World (HNSW) indexing support, parallelization of ivfflat index builds, and improves performance of its distance functions. Post Updated on October 24, 2023 at 10:58PM

Today, AWS CodePipeline announces support for staring a pipeline execution on git tags pushed to repositories in GitHub, GitHub Enterprise Server, GitLab Cloud and Bitbucket Cloud using CodeStarSourceConnection source actions. Customers can now add a “triggers” section in their pipeline definition that has a filter for git tags using include and exclude patterns. When a git tag applied to a commit in the source repository matches the trigger filter, CodePipeline starts a pipeline execution. The CodeStarSourceConnection action clones the commit associated with the git tag. Customers can see the git tag that started the pipeline execution in the console as output variable of the corresponding source action. Customers can use the triggers section using API/CLI/SDK and CDK. Post Updated on October 24, 2023 at 11:11PM

Welcome to Part 4 of our Decoding AI: A government perspective learning series ! In our previous dialogue, we unraveled the complexities of semantic search and generative AI and how they are transforming public governance. In this module, let’s shift gears and focus on the dawn of a new AI era: multimodal sensing. This compelling advancement merges multiple sensory data—think vision, sound, text—into an insightful, unified whole data set. — Siddhartha Chaturvedi, Miri Rodriguez Making sense of different senses In our day-to-day lives, we seldom rely on just one sense to make decisions or understand our environment. Imagine walking into a café; you use your eyes to assess the ambience, ears to catch the vibe through music, and nose to appreciate the aroma of freshly brewed coffee. Similarly, multimodal sensing in AI attempts to capture and process a richer, more nuanced view of the world by combining different types of data. Multimodal sensing: How different types of data can create...

In our recent post , we introduced Microsoft Graph connectors and how they help bring your organization's content into Microsoft Graph and Semantic Index for Copilot to appear across Microsoft 365 experiences, making it easy for users in your organization to easily discover and share content relevant to them while staying in the flow of work. Now, let's go under the hood to better understand the building blocks of a Microsoft Graph connector. Key components of a Microsoft Graph connector First, let’s break things down – Microsoft Graph connectors are responsible for three things: Creating an external connection to your data sources Defining schema for the external content Managing external content that’s imported to Microsoft 365 The following diagram illustrates these responsibilities: An external connection defines the source of external content that you’ll ingest into Microsoft 365. It contains information such as: Name and description Schema, which is a list of pr...

A customer found the byte sequence ba7816bf8f01cfea­414140de5dae2223­b00361a396177a9c­b410ff61f20015ad in the Windows cryptographic libraries. This is the SHA256 hash of the notoriously insecure password abc . (See pages 14 through 16 of the NIST Computer Security Resource Center, Cryptographic Standards and Guidelines, SHA examples document.) Why does the Windows cryptographic library use such a ridiculously weak password, and what is this password used for? While it's true that abc is a horrible password, it's also the case that the Windows cryptographic libraries aren't using it as a password. The value is part of a self-test that the libraries perform to verify that nothing obvious has gone wrong with the standard providers. You can find this hard-coded "well-known SHA256" in the sha256.c module , with the "plaintext" in selftest.c . The values are used by the function Sym­Crypt­Sha256­Self­Test to verify that the algorithm produces the expecte...

We are thrilled to announce the release of Work Item Chart Filtering in Azure DevOps Dashboards. This enhanced filtering streamlines your workflow just a click away. Effortless Filtering for Deeper Insights With this new feature, not only can you hover over a work item chart, but you can also delve deeper into work items by clicking on its segment. The chart will then redirect you to a page where you can view filtered query results. Gone are the days of manual filtering and creating multiple queries to obtain the exact piece of data you need. Dive into your work items and analyze them with a streamlined approach — all in just a few clicks. Simplify Your Workflow : Hover over your work item chart for a quick overview and drill down into specific chart elements for detailed insights by simply clicking on them. Expanded View : Whether your chart is extensive or has lengthy legend names, stay on the dashboard and access an expanded view for a comprehensive look. Contact Us Over th...

Today, AWS re:Post launches Selections , a feature that organizes AWS knowledge for a specific topic, technical issue, or use case. For example, Selections can aggregate content based on a topic such as “ Configure CORS in Amazon S3 ”, a technical issue such as “ Troubleshoot the Amazon S3 403 Access Denied error ”, or a use case such as “ Migrating Windows workloads to AWS .” To date, re:Post has published 10 Selections. re:Post community members with 5,000+ points can publish new Selections.  Post Updated on October 23, 2023 at 06:04PM

Patch Manager, a capability of AWS Systems Manager, now supports instances running Red Hat Enterprise Linux (RHEL) 8.8, MacOS Monterey, and MacOS Ventura. Patch Manager enables you to automatically patch instances with both security-related and other types of updates across your infrastructure for a variety of common operating systems, including Windows Server, Amazon Linux, and Red Hat Enterprise Linux (RHEL). For a full list of supported operating systems, see the Patch Manager prerequisites user guide page .  Post Updated on October 23, 2023 at 07:03PM

Suppose you have a C++ class that is non-copyable and non-movable. This can happen if it has a member which is non-copyable and non-movable, like a std::mutex . But how do you put an object of this class into a map? struct weird { int value; std::mutex mtx; }; std::map<int, weird> table; table.insert({ 1, {} }); // nope table.insert_or_assign(1, weird{}); // nope table.emplace({ 1, {} }); // nope table.emplace(1, weird{}); // nope table.try_emplace(1, weird{}); // nope The problem with the insert method is that it takes a std::pair<int, weird> by value, which means that it cannot move the weird to its final destination. The insert_or_assign , and emplace methods use the parameters to construct a std::pair<int, weird> at the final location, but since we passed a weird{} object as the second parameter, that would require moving the parameter into its final location. The try_emplace method uses its first parameter as a key and the rest of the paramet...

Hello prompt engineers, This week we’re taking a break from code samples to highlight the general availability of Azure AI Content Safety . In this blog series we’ve touched briefly on the using prompt engineering to restrict the types of responses an LLM will provide, such as setting the system prompt to set boundaries on what questions will be answered: Figure 1: System prompt set to "You will answer questions about the speakers and sessions at the droidcon SF conference." However, ensuring a high-quality user experience goes beyond simple guardrails like this. You want your application’s responses to be not only accurate, but also embody your brand’s values and prevent harmful or inappropriate output. With AI capabilities expanding so rapidly, it can be challenging just to keep track of the challenges around ensuring high-quality responses, let alone the steps required to protect against bad output (triggered either by malicious usage or inadvertently). What is Respo...

Today, Amazon QuickSight is announcing the general availability of a native Trino connector that will enable customers to connect to Trino directly from Amazon QuickSight. This launch provides both console and full API support to create, update, edit and delete data sources.  Post Updated on October 20, 2023 at 11:12PM

Today, Amazon QuickSight is announcing the general availability of a native Starburst connector that will enable customers to connect to Starburst directly from Amazon QuickSight. The native Starburst data connector supports connectivity to Starburst Enterprise for on-premises instances and Starburst Galaxy for managed instances.  Post Updated on October 20, 2023 at 11:09PM

If you go browsing through the Win32 API (a common pastime back in the old days), you may run into the Reg­Override­Predef­Key function, which lets a process redirect one of the predefined keys to a separate registry key. What's the idea behind this function? It's explained in the Remarks in the documentation. The idea is that you have a self-registering DLL, and you want to capture the registry changes made by that DLL's Dll­Register­Server function, so that you can take the captured registry changes and add them to your product's main installer. This allows you to simplify your installer to just "Copy these files to these locations, and then set these registry keys to these values." You took the Dll­Register­Server step out of the equation, which speeds up installation and also simplifies auditing. That was the idea, but it pretty much never worked in practice. The trick assumed that all the DLL's Dll­Register­Server did was set some registry keys...

We are retiring older file services versions in October 2023. These older APIs are replaced by Microsoft Graph Drives and DriveItems. We encourage everyone to migrate their applications. After April 2024 , there will be no guarantees that these APIs will continue to function. It is likely you are not using these APIs as they were never documented but please check the list below to see if you are impacted and follow the links for more information on transitioning to modern replacements. _api/web/defaultdocumentlibrary/files This API provides a short-cut method to access the default document library’s files. As an alternative, use Microsoft Graph . https://graph.microsoft.com/v1.0/sites/{site id}/drive Note: “ _api/web/defaultdocumentlibrary ” remains undeprecated and fully supported. _api/beta/files This API is an old implementation of file access within SharePoint and is no longer required. As an alternative, use Microsoft Graph . https://graph.microsoft.com/v1.0/drive/{drive id} h...

本篇翻译于David Ortinau的 Announcing .NET MAUI in .NET 8 Release Candidate 2: More Quality 随着 .NET 8 候选版本 2 (RC2) 中.NET MAUI的发布，我们距离 .NET 8 通用版 (GA) 又近了一步。与 RC1 一样，此版本包含上线许可证，因此您可以在应用程序生产环境中使用它。在此版本中，我们专注于解决在先前版本中出现的问题，并在改进了热重载、可视状态管理器、绑定和应用程序主题的可靠性的同时，恢复了一些性能。 质量改进 除了缩小对回归的关注范围之外，我们还增加了手动测试和自动化测试的覆盖率。.NET 7 和 .NET 8 之间没有重大API 更改，因此您可以期待应用程序升级这方面顺利进行。有关从 .NET 7 升级到 .NET 8 的更多信息，请参考这个简单的 指南 。 此版本的亮点： 性能改进 ： 进行了多项性能改进，包括 Android 上 ActivityExtensions.GetWindowFrame 的性能以及“Setter Specificity”性能。这些优化有助于提高应用程序的流畅性。( #17241 ,  #17364 ,  #17230 ,  #17505 ,  #17545 ). UI 和控件修复 ： 对控件和 UI 元素进行了多项修复和更新，包括 CollectionView、TabBar 可见性、RoundRectangle 边框和 Android 文本对齐相关的修复。有了这些，您的应用程序 UI 在各个平台上更加一致和准确。（#16870、#17240、#17261、 #17353、#17430、#17594）、#17567）。 特定于平台的修复 ： 进行了多项特定于平台的修复，包括拖放功能、选项卡栏外观和特定平台行为，确保了各个平台的用户体验一致。 ( #15748 ,  #16561 ,  #17495 ,  #17041 ,  #17358 . 该版本还包括许多其他错误修复、增强功能和贡献。您可以查看完整的发行说明以了解更多详细信息。 附加信息： .NET MAUI 发行说明 适用于 Android 的 .NET 适用于 iOS 和 Mac 的 ...