Update launching a workflow on Seqera Platform to be a non-blocking action on HPCs
acknowledged
A
Autonomous Bovid
It is possible to encounter request timeouts in Seqera Platform as:
- Launching on the platform is a blocking operation
- There is only 1 connection to an HPC cluster over SSH permitted
- Creating an initial Nextflow job on a cluster may take many seconds which can lead to a timeout (this is cluster specific)
If Seqera can make launching a not blocking action it may help reduce timeouts.
Rob Newman
acknowledged