SSH -> transfer metadata file to slurm
SSH -> check slurm capacity and availability
INT -> find suitable job
SSH -> queue job command include http link for img files
API -> receive guid, job id, timestamp, job queued status
INT -> update asset internally
INT -> keep track of time - after set time has passed -> SSH -> check job in slurm queue -> error handling
API -> receive guid, job id, updated data
INT -> update asset internally -> update ARS
From the integrationservers side:
SSH -> transfer metadata file to slurm SSH -> check slurm capacity and availability INT -> find suitable job SSH -> queue job command include http link for img files API -> receive guid, job id, timestamp, job queued status INT -> update asset internally INT -> keep track of time - after set time has passed -> SSH -> check job in slurm queue -> error handling API -> receive guid, job id, updated data INT -> update asset internally -> update ARS