On a hunch, and seeing in the error message things about Uint64 and Float64 - I wondered if had to do with 32 bit versus 64 bit installations. The ARM installations were 32 bit and x86 installations were 64-bit. So, I downloaded the 32 bit version and installed on my Dell (it’s the master - so to speak). I started up the 32 bit version of Julia and here’s what I got!!
For the Raspberry Pi 2:
julia> addprocs([“julia-user@NODE-RPI2”],dir=“/home/julia-user/julia-0.6.0/bin/”)
1-element Array{Int32,1}:
2
julia> remotecall_fetch(rand, 2, 20)
20-element Array{Float64,1}:
0.438085
0.904493
0.973361
0.679264
0.738671
0.822397
0.175776
0.638466
0.109121
0.234268
0.194145
0.00515439
0.46107
0.0896276
0.367764
0.642768
0.88751
0.835329
0.980536
0.60624
and, the Beaglebone Black:
julia> addprocs([“julia-user@NODE-BBB”],dir=“/home/julia-user/julia-0.6.0/bin/”)
1-element Array{Int32,1}:
3
julia> remotecall_fetch(rand, 3, 20)
20-element Array{Float64,1}:
0.421401
0.697546
0.92379
0.901786
0.754601
0.237044
0.687448
0.223037
0.637141
0.300794
0.526395
0.582173
0.019924
0.584588
0.470795
0.024546
0.995812
0.0406491
0.928998
0.922926