Tools > Allocations... in the Organizer lists each cluster account you have set up and how much of it is used. The figure is for the account as a whole, over all jobs charged to it. It is not what one calculation used; that is the Used row of the calculation's summary.
One row for each machine and account. The accounts come from the machine's profiles (see Allocation profiles); a machine appears if it has an Allocation command, set on the Connection tab of Register Machines, or if its queue manager has a default one (Slurm). Workstations and machines that ECCE does not connect to are not listed.
| Column | Meaning |
|---|---|
| Machine | The machine as registered in ECCE. |
| Account | The account of the profile. |
| Used | The usage, as the cluster reports it. |
| Checked | The time of day the cluster was asked. |
Refresh asks the clusters again. Without it, an answer is kept for 15 minutes. Close closes the window; it can be resized.
4.9 % of 5 000 CPU-h/month used (past 30 days); 3.1 % of 100 GPU-h/month: the output ofprojinfo, as at HPC2N and other NAISS centres. The percentages are of the monthly allocations over the past 30 days.1 235 CPU-h used (no allocation size known): the Slurm default. ECCE asksssharefor the account's usage, which is a total with no allocation size, so no percentage can be given.not recognised: the command ran but ECCE could not read its output. Hold the pointer over the row to see the line of output that names the account. A site's own command can be set in Allocation command; ECCE readsprojinfoand Slurm'ssshareoutput.not checked: not logged in to <machine>: ECCE reads allocations only over a connection that is already open, and never asks for a password for it. Connect to the machine first (for example by launching a job on it), then choose Refresh.checking...: the answer is on its way. The window can be used meanwhile.