Specify your region for compute & storage
This suggestion originally came from a cohort attending a GATK workshop held in Copenhagen, DK.
In Denmark, due to privacy laws around sensitive data, they must keep data & analyses within Denmark. If Terra allowed you to select where your data was kept and where your compute was run, researchers in Denmark and other places with similar limitations would be able to use Terra
Another use case for this is the one that I listed in https://github.com/DataBiosphere/saturn-ui/issues/715:
For our services that run on GCE VMs, we need a way for users to avoid accidental cross-region egress charges. See the GCE network pricing docs.
For AMP PD (and other datasets we want to bring to Terra) we recommend that people store large data (WGS and Transcriptomics) in GCS Regional buckets (and by convention use `us-central1`). This drops storage costs significantly over Multi-Regional buckets especially when data is large.
The one thing you need to be careful about when you do this is that you don't accidentally run a workflow VM in one of the other US regions. We need to help Terra users here. It occurs to me that this applies minimally to:
It would be great if when creating Cloud projects to set the GCE quotas in the non-preferred region to 0, but reducing quotas is currently a manual process involving support.
But it would also be good to have such a setting and for Leonardo and Cromwell under Terra to respect the user's or billing project's settings.
I'm not sure where the status of this feature request is but am replying to bump the thread. This would be lovely for cost savings, to change Terra workspaces to use regional storage and move workspaces into near or coldline storage.
Thanks! I will be sure to bump this internally as well.
We've got the same requests from Australia, due to data privacy reasons we would like the ability to specify the region for compute and storage.
Thanks, Danny Wong. I've added your comment to our internal ticket with the product team.
Same requests I got here in Japan. For data privacy reason, it's necessary to store and analyze data in Japan
Hi Nobuhisa Mizue,
I've added your comment to our internal ticket with the product team and will update this posts with any updates from them!
Please sign in to leave a comment.