Jago Strong-Wright
Jago Strong-Wright
I think everything got cancelled rather than failing but not sure why...
> I think to do this cleanly we might want to follow up with what @vchuravy was doing here #3042 I tried experimenting with this on the OceanBioME tests and...
I may have found another solution: you can limit concurrency of jobs running with buildkite: https://buildkite.com/docs/pipelines/configure/workflows/controlling-concurrency So we can make it so only one instance of the pipeline runs at...
That might be another solution. As it stands each time the pipeline is deployed it downloads julia: https://github.com/CliMA/Oceananigans.jl/blob/eade289dc857396d6d73abde0a4037c1a089d980/.buildkite/pipeline.yml#L20-L24 And then deletes it when its done: https://github.com/CliMA/Oceananigans.jl/blob/eade289dc857396d6d73abde0a4037c1a089d980/.buildkite/pipeline.yml#L862-L866 I guess this was...
For reference, when I fill `u[0, j, k]` instead of `u[1, j, k]` I can just reverse the flow in this cylinder case by changing the sign and swapping the...
I've worked out where my problem is coming from. For the wall-normal velocity: first, we compute and apply the tendencies from 1:N face points, then compute the pressure correction at...
> Can you point me to where in the code this goes down? It happens in this halo fill: https://github.com/CliMA/Oceananigans.jl/blob/45838a57dd5ebc1153c2c827f83cb848d20e4c92/src/Models/NonhydrostaticModels/pressure_correction.jl#L8-L20 after the tendency integration but before the pressure correction
Do you mean when we're computing a diagnostic like vorticity should the kernel include the boundary point? It does seem the simplest would be that anything on a bounded face...
> Where is the halo filling code? Something like: ```julia @inline function _fill_west_open_halo!(j, k, grid, u, bc::PAOBC, loc, clock, model_fields) Δt = clock.last_stage_Δt Δt = ifelse(isinf(Δt), 0, Δt) Δx =...
We only put the no normal gradient matching scheme in the source code which just overwrites the boundary point so this wasn't a problem. I though we weren't going to...