What’s the point: GitLab, VSC, Prometheus, MLflow, AWS

What's the point

Repository management tool GitLab has gotten a set of patches and is now available in versions 11.5.11, 11.6.11, 11.7.12, 11.8.8, and 11.9.9 for GitLab Community Edition and Enterprise Edition as well as 11.10.1 for GitLab CE and EE.

Amongst other things, the latter resolves cluster metrics regression, fixes related merge requests not working with relative URL root, and disables just-in-time Kubernetes resource creation for project level clusters. Versions 11.5.11, 11.6.11, 11.7.12, 11.8.8, and 11.9.9 however have been fitted with a feature flag that should improve performance of accessing git data when using NFS for file storage of Git data.

A small number of fixes only applied to 11.8.8 should get issues concerning the Bitbucket import, health checks, and merge request creation sorted.

Python in Visual Studio Code


Python developers using Visual Studio Code will be interested to learn that the April 2019 release of the project’s Python extension is now available. It comes with a built-in variable explorer and data viewer, as well as some minor changes to the Python Language Server that are meant to lessen the loading time and memory usage when using scientific libraries. Ease of use should be improved with a simplified process to configure the debugger.

Prometheus picks at bugs

Monitoring project Prometheus received a couple of bugfixes in the 2.9.2 release which can be found at GitHub now. The project now makes sure to take a subquery range into account for selection and exhaust every request body before closing it.

Other additions are supposed to fix memory allocation regression and issues in the promtool and remote storage.

MLflow and MS now

Microsoft and Databricks used the opportunity of having a stage at Spark+AI Summit to announce that Microsoft will become an active contributor to Databricks’ MLflow project. This means, amongst other things, that the machine learning project will be natively supported from Azure ML.

In other MLflow news, Databricks has made a fully managed version of the project generally available on AWS and Azure. New additions to the GA version include a way of tracking runs from a sidebar in each Databricks notebook, and a snapshot mechanism that captures the system’s state each time MLflow is used.

AWS gets regional

AWS customers who’d like to get information about regions and services, should now be able to do so via the AWS Systems Manager Parameter Store. Lists of active regions and the available services that go along with it can be queried via the AWS CLI, AWS Tools for Windows PowerShell or the AWS SDKs.

- Advertisement -