fusion

Fusion file system

Tower 22.4 adds official support for the Fusion file system. Fusion is a lightweight client that enables containerized tasks to access data in Amazon S3 (and other object stores in future) using POSIX file access semantics. Depending on your data handling requirements, Fusion 2.0 improves pipeline throughput and/or reduces cloud computing costs. See here for more information on Fusion's features.

Fusion requirements

Fusion file system is designed to work with containerised workloads. Therefore, it requires the use of a container-native platform for the execution of your pipeline. Currently, Fusion is only available in AWS Batch compute environments in Tower.

To enable Fusion in Tower:

• Use Nextflow version 22.10.0 or later. The latest version of Nextflow is used in Tower by default, but a particular version can be specified using NXF_VER in the Nextflow config file field (Advanced options -> Nextflow config file under Pipeline settings on the launch page).

• Enable the Wave containers service during AWS Batch compute environment creation.

• Select Enable Fusion v2 during compute environment creation.

• (Optional) Select Enable fast instance storage to make use of NVMe instance storage to further increase performance.

See the AWS Batch compute environment page for detailed instructions.