After running some parallel simulations with Agent.jl and Distributed.jl, my system is floaded with workers.
I am running parallel simulations with this code:
addprocs(20)
# distribute the model to the worker
@everywhere include("mymodel.jl")
# create a data frame to store the results on each worker
@everywhere adata = [:var1, :var2, :var3, :var4];
# collect everything
results, _ = ensemblerun!(mymodels, agent_step!, dummystep, 252; adata, parallel = true), showprogress=true);
# dismiss the workers
rmprocs(workers())
but after some time I end up with the situation depicted below... what am I missing?
