mining-profit

A tool for tracking historical mining pool profit, to be compared to profit predictions.

CLI Usage

To update with the most recent data, run julia cli.jl -s. For full usage, run julia cli.jl -h

Development Quickstart:

In this directory, use the following procedure to load the package and test it:

In a terminal, run julia to start the repl
Load Pkg with using Pkg
We recommend using Revise: Pkg.add("Revise"); using Revise (if you choose not to do this, you will need to re-import the MiningProfit package after each change)
Activate the package: Pkg.activate(".")
Import MiningProfit: using MiningProfit
Run whatever function you want to test. Consider starting with MiningProfit.updateallstats()

Running in a Production Environment

Motivation for Compilation

There is a lot of speedup to be had from using a precompiled system image. Julia compiles code JIT for the specific parameters you give it, resulting in long wait times for the first run of a given function for a given set of parameter types. While developing Julia code, the Julia community recommends running everything iside a REPL loop to reduce the need to re-compile code. However, this is not a well-suited option when automating processes using julia script. Typically automation is done by calling a script from the command line with cron or similar. Therefore we must either recompile the code on each run (slow) or create a precompiled system image.

To demonstrate the speedup possible, consider the following example for analyzing mining statistics using previously-gathered data. First, the code is run without a pre-compiled system image:

riley-server code/mining-profit ‹main*› » time julia cli.jl -p
  Activating environment at `~/code/mining-profit/Project.toml`
Average Pool profitability:
Dict{Any, Any}("ezil" => 4.311079369316805e-6, "miningpoolhub" => Inf, "flexpool" => 3.2907976861502096e-6)
julia cli.jl -p  27.63s user 1.25s system 100% cpu 28.638 total

Then, after creating a system image with julia build.jl:

riley-server code/mining-profit ‹main*› » time julia --sysimage build/sys_miningprofit.so cli.jl -p
  Activating environment at `~/code/mining-profit/Project.toml`
Average Pool profitability:
Dict{Any, Any}("ezil" => 4.311079369316805e-6, "miningpoolhub" => Inf, "flexpool" => 3.2907976861502096e-6)
julia --sysimage build/sys_miningprofit.so cli.jl -p  7.53s user 0.98s system 103% cpu 8.235 total

This experiment shows nearly 4x speedup! For this reason, using a sysimage is recommended in a production environment.

How to

The ./build.jl script provides an easy way to build a sysimage. Follow these instructions:

Make sure PackageCompiler is installed
In the root of this directory, run julia build.jl to create a precompiled package
From now on, whenever you want to use the program, run julia --sysimage build/sys_miningprofit.so cli.jl <args> to take advantage of improved speeds
If the code is changed (by pulling updates from git or through your own edits) then you will need to re-build the sysimage