A robust, user-level, work-stealing distributed task scheduler.
This tool is still in early development. More details and documentation will follow.
The simplest way to obtain Kueue is by downloading it directly from crates.io. This can be achieved with the following two commands.
Make sure you have a C/C++ compiler installed.
curl https://sh.rustup.rs -sSf | sh
You might need to install OpenSSL headers as well.
cargo install kueue
This will install kueue (the client), kueue_server, and kueue_worker into the bin folder of your Rust installation.
Upon first start of any Kueue binary, a template config file is created at ~/.config/kueue/config.toml.
Make sure that the shared secret in that file is the same on all systems you want to use.
log_level = "info"
server_binds = "0.0.0.0 [::]"
server_name = "ralab29"
server_port = 11236
shared_secret = "keep private!"
Kueue comes with a simple tool named kueue_restart_workers that checks the state of your workers and attempts to restart them if they went down.
To use the tool, add a new block to your config.toml like the following:
[restart_workers]
ssh_user = "klemmefn"
hostnames = """
rax11 rax17 rax19 rax32
ralab04 ralab06 ralab07 ralab08
ralab10 ralab11 ralab13 ralab14
ralab16 ralab18 ralab22 ralab23
ralab24 ralab25 ralab26 ralab27
"""
sleep_minutes_before_recheck = 60
Currently, the tool uses your SSH key to connect to the workers and spawns the worker task using screen. Make sure that screen is installed on your workers and ssh login via key is possible. Then, you can use the tool like this:
eval `ssh-agent -s`
ssh-add ~/.ssh/id_rsa
screen kueue_restart_workers
Keep in mind that kueue_restart_workers is not required for Kueue to work but just a simple tool to make restarting workers simpler. You can also use any other strategy to start and restart your remote workers.