If we want to distribute a build across several build nodes, and want to avoid a "single point of failure", what needs to be considered?
- distribute the build across several build nodes, because some packages take extremely long to build
DBCLIENT="$(which dbclient)" | |
[ -z "$INJECT_USERNAME" ] && INJECT_USERNAME="root" | |
[ -z "$INJECT_PASSWORD" ] && INJECT_PASSWORD="1q2w3e" | |
doremote() { | |
DROPBEAR_PASSWORD="$INJECT_PASSWORD" "$DBCLIENT" -y -y -A -l "$INJECT_USERNAME" "$INJECT_HOST" "$@" | |
} | |
inject_progr() { |
#include <zsdatable.hpp> | |
#include <iostream> | |
int main() { | |
zsdatab::metadata mt(':'); | |
mt += { "k", "w", "v" }; | |
zsdatab::table tab(mt); | |
// insert something into table | |
///... |
ZSXPL -- Zscheile Extensible Programming Language | |
Types: | |
ptr shared_ptr<obj> 0x00 | |
int int64_t 0x01 | |
uint uint64_t 0x02 | |
double double 0x03 | |
stream int32_t(_fd) 0x04 | |
string string 0x10 |
#!/bin/bash | |
export RUST_LOG=debug | |
# source repo @ https://github.com/YZITE/encsess2 | |
# git clone https://github.com/YZITE/encsess2 | |
# cd encsess2 | |
# cargo build --release | |
YZESD2PATH=... # should point at the path where the release binaries reside | |
echo "spawn server ..." |
#!/usr/bin/python3 | |
import filecmp | |
import hashlib | |
import os | |
import shutil | |
import sys | |
args = sys.argv[1:] |
[package] | |
name = "pxlf-client" | |
version = "0.1.0" | |
authors = ["Erik Zscheile <zseri.devel@ytrizja.de>"] | |
edition = "2018" | |
[dependencies] | |
fastrand = "1.4" | |
crossbeam-channel = "0.5" |
#!/usr/bin/python3.9 | |
import argparse | |
import paho.mqtt.publish as mqtt_pub | |
import requests | |
from urllib.parse import urljoin | |
parser = argparse.ArgumentParser() | |
parser.add_argument("host", help="connect to the host (APC) via HTTP") | |
parser.add_argument("user", help="with the given user") |
use prelude::{Service, Box}; | |
pub struct State; | |
impl State { | |
pub fn new() -> Box<dyn Service> { | |
Box::new(Self) | |
} | |
} | |
impl Service for State { |
If we want to distribute a build across several build nodes, and want to avoid a "single point of failure", what needs to be considered?
let nixOp = nixBlti.nixOp; | |
let nixBltiRT = nixBlti.initRtDep(nixRt); | |
let nixInScope = nixBlti.mkScopeWith(); | |
return async (nixBound) => { | |
nixBound = await nixBound; | |
let nix__lib = nixOp._lambdaA2chk(nixBound, "lib"); | |
return (async (nixInScope) => | |
(async (nixInScope) => | |
(async (nixInScope) => | |
nixBlti.PLazy.from(async () => { |