Rafael Ponte rponte

Race conditions, locking e bancos de dados relacionais

Pretendo apresentar um código com uma lógica de negócio simples que funciona bem na perspectiva do negócio mas que quebra miseravelmente em ambientes minimamente concorrentes. Para isso, vou demonstrar através de testes de integração como identificar o Race Condition no código e principalmente como resolvê-lo através de mecanismos de locking e Isolation Level do seu banco de dados;

O que vamos discutir

Por que estudamos esse assunto mais a fundo?
- 1.1. Bootcamp Seniors
- 1.2. Handora: Curso CRUD e Atualizações Concorrentes com JPA/Hibernate - by Jordi Silva

1.3. A importância do treino: o que eu levei alguns bons anos para aprender o Jordi levou 6 meses

title
Elements of Messaging and Eventing Platforms

This document provides a brief overview of the essential elements of a messaging and eventing platform and how they relate to each other.

Message and Event Broker Categories

Thread Pools

Thread pools on the JVM should usually be divided into the following three categories:

CPU-bound
Blocking IO
Non-blocking IO polling

Each of these categories has a different optimal configuration and usage pattern.

Things I believe

This is a collection of the things I believe about software development. I have worked for years building backend and data processing systems, so read the below within that context.

Agree? Disagree? Feel free to let me know at @JanStette. See also my blog at www.janvsmachine.net.

Fundamentals

Keep it simple, stupid. You ain't gonna need it.

Distributed Transactions and why you should avoid them

Modern technologies won't support it (RabbitMQ, Kafka, etc.);
This is a form of using Inter-Process Communication in a synchronized way and this reduces availability;
All participants of the distributed transaction need to be avaiable for a distributed commit, again: reduces availability.

Implementing business transactions that span multiple services is not straightforward. Distributed transactions are best avoided because of the CAP theorem. Moreover, many modern (NoSQL) databases don’t support them. The best solution is to use the Saga Pattern.

[...]

	import org.slf4j.LoggerFactory
	import org.springframework.jdbc.core.JdbcTemplate
	import org.springframework.stereotype.Component
	import org.springframework.transaction.annotation.Propagation
	import org.springframework.transaction.annotation.Transactional
	import java.time.Duration

	interface LockManager {
	fun <T> tryWithLock(key: Long, timeout: Duration, function: () -> T): T
	}

	--
	-- Trying to buy a new ticket for an event
	--
	BEGIN;
	UPDATE events e
	SET updated_at = now()
	WHERE e.id = :event_id
	AND e.max_tickets > (SELECT count(*) -- (does this logic work with READ_COMMITTED??)
	FROM tickets t
	WHERE t.event_id = e.id);

	#!/usr/bin/env sh
	#
	# Simple script that prints out the top N most-used words in a text from standard input.
	#
	# Inspired by https://buttondown.email/hillelwayne/archive/donald-knuth-was-framed/. The short
	# linux script is first shown at https://www.cs.tufts.edu/~nr/cs257/archive/don-knuth/pearls-2.pdf

	topw() {
	tr -cs A-Za-z '\n' \| tr A-Z a-z \| sort \| uniq -c \| sort -rn \| sed "$N"q
	}


	-- Let's say you have a table full of work:

	CREATE TABLE tasks (
	id UUID PRIMARY KEY NOT NULL DEFAULT gen_random_uuid(),
	status TEXT NOT NULL DEFAULT 'pending',
	payload JSON NOT NULL, -- or just have meaningful columns!
	created_at TIMESTAMP NOT NULL DEFAULT NOW()
	);

	#!/bin/bash -eu
	if [ -z "${LIB_ROBUST_BASH_SH:-}" ]; then
	LIB_ROBUST_BASH_SH=included

	function error {
	echo "Error: ${1:-}"
	}

	# Check to see that we have a required binary on the path
	function require_binary {