Describe the bug
An error seems to occur when running ww3_multi in the mww3_test_05/./work_ST6_PR2_UQ_OMP regression test at current head of develop (761cf79) on our UKMO Cray GNU. I had not experienced this when running this on older commits e.g (c3b0d04)
To Reproduce
Run the mww3_test_05 regression test and check output logs
Expected behavior
The test to successfully produce ´files for ww3.68060600, ww3.68060601, ww3.68060602, ww3.68060603, ww3.68060604, ww3.68060605, ww3.68060606
Screenshots
Here's a comparison of the ww3_multi.out between a test working as expected (left) and one with the error seen (right):
Additional context
I observed this problem when running regression tests for #1613. Counterintuitively, running on the head of that branch (d71f926) did not cause the issue to appear but running on head of develop as above did - despite the only code difference being a change to NetCDF metadata defaults. This was repeated upon retesting, hence I consider it an intermittent issue.
As above, this is the first time I recall encountering the error.
I apologise in advance that I am not familiar with the particularly regression test and I have not dug into investigating the bug particularly deeply but I'll be happy to provide any further info or clarifying context.
Describe the bug
An error seems to occur when running ww3_multi in the mww3_test_05/./work_ST6_PR2_UQ_OMP regression test at current head of develop (761cf79) on our UKMO Cray GNU. I had not experienced this when running this on older commits e.g (c3b0d04)
To Reproduce
Run the mww3_test_05 regression test and check output logs
Expected behavior
The test to successfully produce ´files for ww3.68060600, ww3.68060601, ww3.68060602, ww3.68060603, ww3.68060604, ww3.68060605, ww3.68060606
Screenshots
Here's a comparison of the ww3_multi.out between a test working as expected (left) and one with the error seen (right):
Additional context
I observed this problem when running regression tests for #1613. Counterintuitively, running on the head of that branch (d71f926) did not cause the issue to appear but running on head of develop as above did - despite the only code difference being a change to NetCDF metadata defaults. This was repeated upon retesting, hence I consider it an intermittent issue.
As above, this is the first time I recall encountering the error.
I apologise in advance that I am not familiar with the particularly regression test and I have not dug into investigating the bug particularly deeply but I'll be happy to provide any further info or clarifying context.