What’s New?

The API path for managing technologies has been changed.

External technologies has been removed from Saagie Official catalog.

User identities and access to Object Storage and PostgreSQL are now manageable via OVH admin interface.

Saagie now supports Kubernetes 1.29.x.

The Intercom chatbot is being replaced by a dedicated Saagie Helper.

Spark jobs can no longer be scheduled under one hour.

You can now rerun a previous pipeline instance.

GPU resource management on jobs has been added.

When you schedule your job or pipeline, a new option allows you to prevent more than one instance of it from running at the same time.

Saagie now supports Kubernetes 1.28.x.

The job execution timeline has been enhanced to show more details.

When you push your pipeline using the Saagie CI/CD process, a link to your pipeline’s source code is now added to the pipeline’s graph.

You can now view, copy, import, and export your pipeline’s code in its Edit mode.

You can now create environment variables at the app level.

Saagie now supports Kubernetes 1.26.x and 1.27.x.

Saagie is now compatible with the cgroup v2 version of the Linux cgroup API.

The upgrade procedure for this new version of Saagie is a bit different.

You can now move a pipeline from one project to another.

The function for duplicating a pipeline has been improved.

Pipelines now have an alias in addition to their name, just like jobs.

Management of the storage space size has been improved.

You can now also duplicate your app storage spaces.

When you push your job using a CI/CD process, a link to your job’s source code is now added to your job package.

You can now log in to Saagie using a single sign-on (SSO) process.

Find the Saagie CI/CD GitHub Action on the GitHub marketplace. Its aim is to make it easier to set up the CI/CD process on your Saagie platform.

Storage space management has been improved. You can now move a storage space to another project, as well as increase the capacity of a storage space.

It is now possible to duplicate a pipeline.

A new capacity planning dashboard has been added to the Saagie Usage Monitoring (SUM) add-on.

Learn how to resume a pipeline execution from the point at which it stopped if it failed.

Learn how to do machine learning pipelines using BigQuery with Python in Saagie.

Learn how to push models in the Hugging Face’s Model Hub with Saagie.

Display a global view of your cluster consumption with Saagie Resources Monitoring.

The function for moving a job from one project to another has been improved.

You can now see and follow the details of your job execution time.

You can delete pipeline instances and versions.

You can delete job instances based on date criteria.

You can also delete job versions based on tag criteria.

You can integrate your Saagie projects into a CI/CD pipeline.

You can generate a job description with ChatGPT.

You can configure your Jupyter Notebook app for use with generative AI.

You can configure your VS Code app too for use with AI Pair Programmers.

A new add-on named Saagie Hugging Face Model Server has been released. It helps you deploy and predict Hugging Face deep learning models for text classification.

Another add-on named Saagie Code Search has been released too. It helps you search and retrieve Python code snippets from existing codebase.

Saagie now supports Kubernetes 1.25.x.

You can now delete job instances and versions.

You can now duplicate a job.

Default values for CPU and RAM resources have been defined for all technologies, except for external technologies. In addition, these resource capacities are now enabled by default when creating jobs and apps with the predefined default values.

A patch has been released to handle ambiguous floating values in the attributes of the technology’s metadata.yaml files.

Job execution lasting more than 15 minutes now end with an appropriate status, instead of Unknown.

A pagination has been implemented on the app History page to improve page loading fluidity.

New elements have been created to monitor resource consumption of pipelines and your cluster.

Pipeline functionality has been enhanced to include more advanced orchestration logic, such as conditions on environment variables and job status.

Saagie now supports Google Cloud Platform (GCP).

New elements to monitor resource consumption have been created.

A new add-on, called Saagie Usage Monitoring, can be deployed as an app inside projects.

Pipeline functionality has been enhanced to allow context propagation between jobs in a pipeline.

Saagie now supports Kubernetes 1.23.x and 1.24.x.

Saagie will now be installed with a ready-to-use example project, which goal is to propose an intelligent learning pipeline able to detect feelings on movie reviews.

The product version naming pattern has changed.

The management of apps in Saagie has been overhauled.

The external technologies can now be orchestrated in Saagie.

Information about the product version are now available.

The Overview page of jobs and pipelines has been redesigned.

The user experience in pipelines has been improved.

Breaking changes in the Saagie API behavior have been made.

Multi-context management is now available for app technologies in Saagie’s official Technology Catalog.

The user experience in pipelines has been improved.

You can create conditional pipelines to run multiple jobs in parallel.

Resource allocation quotas can now be defined for each job to control their resource consumption.

You can now select time zones when configuring scheduled runs for jobs and pipelines.

You can now physically delete projects, jobs, pipelines, environment variables, and Docker credentials from their platforms.

Fixes have been applied to all Spark runtime contexts available in Saagie repositories.

Job, pipeline, and app status updates are now automatic.

A sample project is available when installing a new platform.

The default email of the platform sender can be customized.

You can now configure the persistent storage size when creating apps.

You can now monitor various Saagie metrics with Prometheus.

The pattern for the platform URLs has changed.

A new page has been added: the Activity page. It allows you to get a quick overview of the latest job and pipeline instances executed.

The GPU option is now available on jobs.

The navigation within a project has been improved.

The technology details are visible at a glance when creating a project.