MaximumCapacityReached, but scheduler insists in submitting to resource

[GoogleCodeExporter]

1. submit to an OpenStack cloud,
2. set max_vm to a limit lower than the quota
3. submit more jobs than the quota can handle
4. at some point, you will staert seeing submission errors: "informing 
scheduler: MaximumCapacityReached: Already running the maximum number of VMs on 
resource"

...but the scheduler is apparently never informed, as the errors accumulate.

Original issue reported on code.google.com by riccardo.murri@gmail.com on 24 Apr 2014 at 3:38

[GoogleCodeExporter]

Original comment by riccardo.murri@gmail.com on 24 Apr 2014 at 4:07

[GoogleCodeExporter]

. Does the 'error accumulate' over the same submit cycle (meaning that the 
scheduler still tries to submit another Application to the same resource) ? 

. What would be the expected behavior for the scheduler ?

S.

Original comment by sergio.m...@gmail.com on 28 Apr 2014 at 2:05

[GoogleCodeExporter]

| . Does the 'error accumulate' over the same submit cycle (meaning that the
| scheduler still tries to submit another Application to the same resource) ?

Yes.

| . What would be the expected behavior for the scheduler ?

Stop submitting to that resource for the current cycle.

Original comment by riccardo.murri@gmail.com on 28 Apr 2014 at 2:30