Data Resources

Acacia is our disk-based storage system, powered by Dell, providing 60PB of high-speed object storage for hosting research data. This multi-tiered cluster separates different types of data to improve data availability. Acacia has been built around Ceph in-house. It is highly scalable, stable and resilient. Acacia is a significant technological change from our past storage infrastructure. With it, we have embraced object storage technologies which make for a more flexible ecosystem where research groups can more easily share and access data. We have also moved to using the open-standard S3 interface for access.

If you already have an active, approved Pawsey project (Supercomputing, Nimbus or Visualisation) then you are automatically eligible for a 1TB project storage and a 100GB private storage allocation on Acacia.

To learn how to start using your warm tier allocation, please read the Acacia - User Guide.

Banksia is an in-place upgrade to our previous tape-based system, which consisted of two Spectralogic T-Finity tape libraries, each with a current storage capacity of 74 PB - mirrored for greater data protection. Data was migrated from the old Pawsey HSM to the new system, via an in-place data migration, whereby the old format tapes continue to be read by Versity software. This system was supplied by Xenon Australia and is managed by the ScoutAM platform from Versity operating on Intel servers. It has a 5 PB front-end DDN hybrid storage cache operating with ScoutFS to take advantage of the centre’s new 100 Gb network infrastructure. Both tape libraries have received an upgraded robotics control module, and are connected by new IBM fibrechannel switches. They each contain 32 IBM® TS1150 high performance enterprise tape library drives and 2 IBM® TS1160 next generation drives accessing over 10,000 media slots.

Similar to the warm tier, we have shifted towards adopting more open standards by migrating to an open tape format, and a new event broker (Kafka) to suit event-driven workflows.

Banksia remains the home of our existing data projects, including the multi-PB SKA radio astronomy data, and is available via a managed storage application.

To learn how to start using your cool tier allocation, please read the Mediaflux Guide.