Issue with interactive and default threadpools in Julia 1.9

[simsurace]

I want to keep low latency tasks on the interactive thread pool and spawn heavier tasks on the default threadpool. However, this still doesn't seem to work with the most recent release candidate:

The following is a MWE server script. Run with julia +beta -t1,1

using Dates, HTTP

const ROUTER = HTTP.Router()

function low_latency(req)
    @info "$(now())> $(Threads.threadpool()) tid: $(Threads.threadid())"
    return "Done with low latency task"
end
HTTP.register!(ROUTER, "GET", "/low", low_latency)

function high_latency(req)
    Threads.@spawn :default begin
        s = sum(randn(10^8))
        @info "$(now())> $(Threads.threadpool()) tid: $(Threads.threadid()), sum = $s"
        return "Done with high latency task"
    end
end
HTTP.register!(ROUTER, "GET", "/high", high_latency)

function requestHandler(req)
    resp = ROUTER(req)
    return HTTP.Response(200, resp)
end

HTTP.serve(requestHandler, "0.0.0.0", 7777)

Calling the /low route works fine, but on calling the /high route, I get:

[ Info: 2023-05-02T16:25:18.769> default tid: 2, sum = 3245.4070974669858
┌ Error: handle_connection handler error
│   exception =
│    MethodError: no method matching length(::Task)
│    
│    Closest candidates are:
│      length(::Union{Base.KeySet, Base.ValueIterator})
│       @ Base abstractdict.jl:58
│      length(::Union{LinearAlgebra.Adjoint{T, S}, LinearAlgebra.Transpose{T, S}} where {T, S})
│       @ LinearAlgebra ~/.julia/juliaup/julia-1.9.0-rc3+0.x64.linux.gnu/share/julia/stdlib/v1.9/LinearAlgebra/src/adjtrans.jl:295
│      length(::Union{SparseArrays.FixedSparseVector{Tv, Ti}, SparseArrays.SparseVector{Tv, Ti}} where {Tv, Ti})
│       @ SparseArrays ~/.julia/juliaup/julia-1.9.0-rc3+0.x64.linux.gnu/share/julia/stdlib/v1.9/SparseArrays/src/sparsevector.jl:95
│      ...
│    
│    Stacktrace:
│      [1] mkheaders(h::Task)
│        @ HTTP.Messages ~/.julia/packages/HTTP/qGpdS/src/Messages.jl:209
│      [2] HTTP.Messages.Response(status::Int64, headers::Task, body::Vector{UInt8}; version::HTTP.Strings.HTTPVersion, request::Nothing)
│        @ HTTP.Messages ~/.julia/packages/HTTP/qGpdS/src/Messages.jl:108
│      [3] HTTP.Messages.Response(status::Int64, headers::Task; body::Vector{UInt8}, request::Nothing)
│        @ HTTP.Messages ~/.julia/packages/HTTP/qGpdS/src/Messages.jl:112
│      [4] HTTP.Messages.Response(status::Int64, headers::Task)
│        @ HTTP.Messages ~/.julia/packages/HTTP/qGpdS/src/Messages.jl:112
│      [5] requestHandler(req::HTTP.Messages.Request)
│        @ Main ~/Documents/projects/n/julia-ai-service/test.jl:22
│      [6] (::HTTP.Handlers.var"#1#2"{typeof(requestHandler)})(stream::HTTP.Streams.Stream{HTTP.Messages.Request, HTTP.Connections.Connection{Sockets.TCPSocket}})
│        @ HTTP.Handlers ~/.julia/packages/HTTP/qGpdS/src/Handlers.jl:58
│      [7] #invokelatest#2
│        @ ./essentials.jl:816 [inlined]
│      [8] invokelatest
│        @ ./essentials.jl:813 [inlined]
│      [9] handle_connection(f::Function, c::HTTP.Connections.Connection{Sockets.TCPSocket}, listener::HTTP.Servers.Listener{Nothing, Sockets.TCPServer}, readtimeout::Int64, access_log::Nothing)
│        @ HTTP.Servers ~/.julia/packages/HTTP/qGpdS/src/Servers.jl:447
│     [10] macro expansion
│        @ ~/.julia/packages/HTTP/qGpdS/src/Servers.jl:385 [inlined]
│     [11] (::HTTP.Servers.var"#16#17"{HTTP.Handlers.var"#1#2"{typeof(requestHandler)}, HTTP.Servers.Listener{Nothing, Sockets.TCPServer}, Set{HTTP.Connections.Connection}, Int64, Nothing, Base.Semaphore, HTTP.Connections.Connection{Sockets.TCPSocket}})()
│        @ HTTP.Servers ./task.jl:514
└ @ HTTP.Servers ~/.julia/packages/HTTP/qGpdS/src/Servers.jl:461

Is this user error or a bug?

Versions

• Julia 1.9.0-rc3
• HTTP.jl 1.8.0
• MbedTLS.jl 1.1.7

[quinnj]

Request handlers must be of the form f(::Request) -> Response), and you're violating that for the high latency; instead you want:

function high_latency(req)
    fetch(Threads.@spawn :default begin
        s = sum(randn(10^8))
        @info "$(now())> $(Threads.threadpool()) tid: $(Threads.threadid()), sum = $s"
        return "Done with high latency task"
    end)
end

[simsurace]

Oh my bad, thanks a lot!