What’s the point: Azure DevOps roadmap, Spring Boot, Skopeo, Drill in need, TriggerMesh, and RedisAI

What’s the point: Azure DevOps roadmap, Spring Boot, Skopeo, Drill in need, TriggerMesh, and RedisAI

The DevOps section of Microsoft’s Azure team has updated the roadmap for the service formerly known as Team Foundation Server. Amongst other things, Azure DevOps is meant to see the introduction of service tags before the end of Q2. The tags represent a group of IP prefixes which can be used for IP whitelisting, for instance.

Pipeline users will surely appreciate the plans to move elastic self-hosted agents into public preview and multi-stage pipelines into general availability. Apart from that, the service will soon allow multi-repo triggers, so that pipeline YAMLs don’t have to reside within the same repo as the microservices they manage. 

Admins using Azure Boards meanwhile will get the capability to restrict the transition from one state to another on a work item type, and configure policies to prevent project and team admins from inviting additional users. New events to keep track of alternate credential usage and token lifecycle management activities promise an extra layer of auditability.

Spring Boot 2.3 starts taking app’s pulse

Java-microservice framework Spring Boot is now available in version 2.3. Those looking for interesting features will mostly find updated dependencies and improved support for Docker and Java 14. However, v2.3 also supports graceful shutdown on Jetty, Reactor Netty, Tomcat, and Undertow as well as with reactive and Servlet-based web applications, and probes applications for liveness and readiness, so there’s plenty of reason to give it a go.

Skopeo follows Podman and Buildah over the 1.0 line

Command line utility Skopeo has finally passed the 1.0 version mark, making the project essentially ready for production. To make this milestone, the team has mostly been busy stabilising the project with fixes and tests. However, Skopeo 1.0 also features login and logout commands meant to help users access container registries, and a security policy.

Skopeo can be used to copy container images between different types of container storage, as well as delete and inspect them. Compared to other projects it doesn’t require a daemon to be running in order to perform operations, and users don’t need root access for most things, which comes in handy in an enterprise environment. It is often used in combination with container engine Podman and build tool Buildah, all of which are Red Hat-bred projects.

Apache Drill: Reviewers apply within

Using Apache Drill in a company project? Time to help the team out! The framework for data-intensive distributed applications has lost its primary backer, HPE. The company will still be “contributing minimal bug fixes”, adding new functionality or features however is now up to those who still feel the project has value to them.

To keep the project going, Drill now needs volunteers to review code commits. “Under the Apache rules, to commit code to the code base it must be reviewed and voted upon. While we have committers, we do not currently have enough code reviewers to review new commits.” wrote Charles Givre, chair of the Drill project management committee, last week.

TriggerMesh EverBridge starts beta program

Cloud-native integration platform TriggerMesh has made the jump into the messy world of legacy code by pushing its EveryBridge event bus into beta. The project is meant to help friends of hybrid architectures combine serverless offerings across cloud platforms with custom applications and more established event sources such as VM vSphere, and Solace. 

Redis makes AI and serverless products generally available
Redis Labs, the company behind in-memory data structure store, has pushed out version 1.0 of its AI serving engine RedisAI and serverless offering RedisGears. While the latter can be used for tasks such as real-time data and event processing, with operations across multiple data structures, RedisAI was built to “simplify a scalable deployment of AI models” and minimise the amount of time needed for inference processes.