Skip to content

Release notes for version 21.02#

Highlights#

Tower 21.02.x introduces support for:

  • Add Azure Batch provider
  • Add Altair PBS pro provider
  • Add sessionId to workflows search-box criteria
  • Add support for multiple GLS zones
  • Add Grid provider head job options
  • Add support for AWS Batch cost percentage
  • Add Azure Batch Forge
  • Add support for Grid Engine batch scheduler
  • Add Kubernetes service pod
  • Add support for Tower license
  • Improve detection of NF config profiles #1074
  • Fix issue on work dir path composition with ending slash
  • Fix issue when retrieving non-existing file via SSH/SCP
  • Fix issue resolving non-canonical GitHub/Gitlab project name #353
  • Fix issue with AWS Batch allocation strategy #931
  • Fix job phantom unknown status
  • Fix Prevent requeue mail with invalid addresses
  • Fix issue on creating AWS CE with manual config
  • Update backend base image to corretto:11.0.10
  • Bump NF 21.03.0-edge
  • Upgrade to Angular 11
  • Use Kubernetes Java-client 10.0.1
  • Upgrade Nextflow runtime to 21.03.0-edge

Updating Tower deployment from version 20.12.x to 21.02.x#

License key#

As of this version, a license key must be provided to enable the Tower deployment feature. The license key should be specified using the configuration variable TOWER_LICENSE.

Warning

If you don't have a license key, contact sales@seqera.io.

Compute environments#

The Tower compute environments that are available to users must be specified in the Tower configuration.

The following ids options are available:

  • awsbatch-platform: AWS Batch cloud compute service
  • gls-platform: Google LifeSciences cloud compute service
  • azbatch-platform: Azure Batch cloud compute service
  • lsf-platform: IBM LSF batch scheduler
  • slurm-platform: Slurm batch scheduler
  • altair-platform: Altair PBS Pro batch scheduler
  • univa-platform: (Univa/Sun) GridEngine
  • k8s-platform: Kubernetes compute platform
  • eks-platform: AWS EKS compute platform
  • gke-platform: Google GKE compute platform

Choose one or more of these platform ids and append to your current MICRONAUT_ENVIRONMENTS variable, separating them via a comma.

Database schema#

This Tower version requires a database schema update. Follow these steps to update your DB instance and the Tower installation.

Kubernetes based deployment#

1. Update the Tower container images in the Kubernetes manifest yaml files to:

1
2
195996028523.dkr.ecr.eu-west-1.amazonaws.com/nf-tower-enterprise/backend:v21.02.5
195996028523.dkr.ecr.eu-west-1.amazonaws.com/nf-tower-enterprise/frontend:v21.02.5

If you are using AWS Batch with a custom launcher job definition you need to update it to use the following container image (please refer the configuration section for details):

1
public.ecr.aws/seqera-labs/tower/nf-launcher:21.04.0-edge

Refer to the manifests included in the Kubernetes section for details.

2. Update the Tower cron service using the following:

1
kubectl apply -f tower-cron.yml

Note

This task will automatically run the Tower database schema update tool.

3. Update the Tower backend and frontend services using the following command:

1
kubectl apply -f tower-svc.yml

Custom deployment script#

1. Pull or update the Tower container images references in your deployment script(s) to:

1
2
195996028523.dkr.ecr.eu-west-1.amazonaws.com/nf-tower-enterprise/backend:v21.02.5
195996028523.dkr.ecr.eu-west-1.amazonaws.com/nf-tower-enterprise/frontend:v21.02.5

If you are using AWS Batch with a custom launcher job definition you need to update it to use the following container image (please refer the configuration section for details):

1
public.ecr.aws/seqera-labs/tower/nf-launcher:21.04.0-edge

2. Update the Tower database schema by running the /migrate-db.sh provided in the backend container.

Note

Make sure to include the identical environment as used in the normal backend execution.

3. Once the schema update completes, deploy Tower following your usual procedure.

Back to top