Seqera Platform Enterprise version 25.3 introduces a series of new features and enhancements, including custom user roles, Studios integration with Git providers, and pipeline versioning.
Feature updates and improvements
Custom roles and fine-grained authorization (RBAC)
- Added support for custom roles with fine-grained permissions, to define specific access controls beyond the predefined Owner, Admin, Maintain, Launch, View, and Connect roles.
Studios
- Updated the Add Studio page to a single-step form.
- Added support for Azure single VM compute environments.
- Added support for loading Studio configuration from private and public Git repositories.
- Added ability to clone content from a Git repository to a Studio at startup time.
- Added ability to create Studios from Docker images, built from a repository Dockerfile using Wave.
- Added ability to stop a Studio session that is stuck starting without needing AWS console access.
- Added Save as data link action when mounting data to a Studio.
- Added support for custom environment variables to Studio configuration.
Pipelines
Datasets
- Refactored datasets list and detail views with additional metadata and improved version control.
- Added ability to dynamically show/hide datasets for all users in the workspace. Hidden datasets do not consume workspace quota.
- Removed dataset quota limits.
- Added ability to sort dataset list by name, recently updated, and recently used.
- Added support for labels.
- Enhanced filtering of all dataset attributes (visibility, label, username,
createdAfter, createdBefore, updatedAfter, updatedBefore, usedAfter, usedBefore).
- Added one-click view of all pipeline runs using a specific dataset, adding direct traceability between datasets and workflow runs.
Data Explorer
- Added support for S3-compatible object storage providers. To use additional object storage providers, update the unprivileged CSP policy.
- Added support for LakeFS Cloud in Data Explorer by enabling image preview and file upload functionality when using LakeFS Cloud as an S3-compatible storage provider.
- Added CSP config for OCI Object Storage, Cloudflare R2, and LakeFS Cloud.
Compute
- Configured Nextflow to automatically terminate unschedulable jobs on AWS Batch compute environments by setting
aws.batch.terminateUnschedulableJobs to true by default. This prevents jobs that cannot be scheduled (due to invalid resource requests or configuration errors) from indefinitely blocking the job queue.
- Implemented safeguards to prevent unschedulable jobs from blocking other jobs in the queue. When jobs cannot be scheduled due to resource constraints or configuration errors, they are handled separately to ensure other valid jobs can proceed.
- Added the AWS
g6e instance family (NVIDIA L40S GPUs) to the NVMe disk support list, enabling high-performance instance storage for GPU-accelerated workloads.
- Added
f2.6xlarge, f2.12xlarge, f2.48xlarge instance types.
- Extended Google Cloud config with selectable Advanced options -
machineType, image, armEnabled.
Nextflow launcher image
If you host your nf-launcher container image on a private image registry, copy the nf-launcher image to your private registry. Then update your tower.env with the following environment variable:
TOWER_LAUNCH_CONTAINER=<FULL_PATH_TO_YOUR_PRIVATE_IMAGE>
If you're using AWS Batch, you will need to configure a custom job-definition and populate the TOWER_LAUNCH_CONTAINER with the job-definition name instead.
General
- Bumped Nextflow launcher to
nf-launcher:j21-25.04.8.
- Replaced the top navigation with a sidebar in Platform.
- Updated Platform behavior to redirect to first available sidenav item instead of hardcoded launchpad.
- Added usage statistics to the Datasets API, allowing you to view the number of pipeline runs using each dataset and when it was last used.
- Added metrics to confirm whether a pipeline is running GPU resources.
- Added better support for treating Nextflow config preferentially, as mono-repo configurations now read from the same directory as the configured main script.
- Added support for BitBucket API tokens to ensure that Bitbucket can still be used as an API provider. See the announcement for more information: https://bitbucket.org/account/settings/app-passwords/.
Bug fixes
Studios
- Fixed Studios AWS cloud private ECR pull issues.
- Fixed handling of Studios tokens without user ID claim.
- Fixed trigger YAML validation on Studios config patch.
Pipelines
- Fixed pipeline list not updating when changing workspace.
- Fixed pipeline config reading from same directory as main script.
- Fixed
describePipeline to properly handle avatar files.
Workflows
- Prevented incorrect
workdir patching when resuming workflow with compute environment changes.
- Added proper permission checks for workflow star functionality.
Compute environments
- Fixed transactional event listener annotation for compute operations.
- Improved Connect parameter handling for compute environments.
- Fixed error when saving long container names.
UI/UX
- Fixed MultiQC plots displaying as grey boxes.
- Fixed badge cropping issues in Platform UI.
- Fixed infinite scroll in optimization profile modal.
- Forced perfect circle avatar shape for consistent rendering.
Permissions and roles
- Removed blanket
ListOrganizationSettings check, use individual permissions instead.
General
- Fixed OIDC API handling in authentication flows.
Fusion
See the Fusion changelog for more details.
Wave
See the Wave changelog for more details.
Nextflow
See the Nextflow migration notes page for changelog details.
MultiQC
See the MultiQC changelog for details.
General warnings
Starting from version 26.1, the front end image running as root user will be deprecated. We recommend starting to switch to the root-less image (also known as "unprivileged" image) during this upgrade. See Kubernetes and docker-compose deployment documentation for instructions.
Upgrade steps
This release maintains backward compatibility with version 25.2.x.
- Make a backup of your Platform database prior to upgrade.
- If you are upgrading from a version prior to 25.1, complete all intermediate major version upgrades before upgrading to 25.3.
- Ensure that no pipelines are in a running state during this upgrade as active run data may be lost.
See Upgrade deployment for installation guidance.
