discourse/migrations/tooling/scripts/benchmarks/write.rb
Gerhard Schlager 06b32204c0
MT: Split the migrations tooling into separate gems (#40492)
Previously, the migrations tooling was a single flat `migrations/` tree,
autoloaded by one global Zeitwerk loader and driven by a Thor CLI, so each
planned next step had nowhere clean to land.

This change splits it into four `path:`-referenced gems — `migrations-core`,
`migrations-tooling`, `migrations-converters`, and `migrations-importer` —
served by a single Samovar-based `disco` binary, without rewriting any domain
logic.

### Why now

The DSL refactor that replaced the IntermediateDB YAML config just landed,
which is the cheapest moment to do this. Everything queued behind it — column
coverage verification, the `discourse-migrations` validation plugin, the
transformer framework, and private converter isolation — either has nowhere
clean to land in the flat tree or would have to be retrofitted into a gem
layout later. Doing the split now, while it's still a pure move (suite green,
no domain logic touched), is far cheaper than after another round of features
has built on the flat layout.

### What changes

- **Four gems under `migrations/`**, all `path:`-referenced from the root
  `Gemfile` (nothing is published to RubyGems): `core` (CLI framework, UI, DB
  infrastructure, IntermediateDB, and the conversion framework), `tooling`
  (schema DSL and `schema` commands), `converters` (implementations and source
  adapters), and `importer` (row and uploads import).
- **A single CLI binary:** `migrations/bin/cli` (Thor) becomes `disco`
  (Samovar), with each gem registering its own commands. Same surface —
  `convert`, `import`, `upload`, `schema generate|validate|…` — and Rails is
  still booted lazily.
- **Isolated test suites:** each gem runs its own no-Rails specs in a new CI
  job, while the existing job keeps running the Rails-integration specs.
2026-06-02 22:20:03 +02:00

186 lines
3.9 KiB
Ruby
Executable file
Vendored

#!/usr/bin/env ruby
# frozen_string_literal: true
require "bundler/inline"
gemfile(true) do
source "https://rubygems.org"
gem "extralite-bundle"
end
require "etc"
require "extralite"
require "tempfile"
SQL_TABLE = <<~SQL
CREATE TABLE users (
id INTEGER,
name TEXT,
email TEXT,
created_at DATETIME
)
SQL
SQL_INSERT = "INSERT INTO users VALUES (?, ?, ?, ?)"
USER = [1, "John", "john@example.com", "2023-12-29T11:10:04Z"]
ROW_COUNT = Etc.nprocessors * 200_000
def create_extralite_db(path, initialize: false)
db = Extralite::Database.new(path)
db.pragma(
busy_timeout: 60_000, # 60 seconds
journal_mode: "wal",
synchronous: "off",
)
db.execute(SQL_TABLE) if initialize
db
end
def with_db_path
tempfile = Tempfile.new
db = create_extralite_db(tempfile.path, initialize: true)
db.close
yield tempfile.path
db = create_extralite_db(tempfile.path)
row_count = db.query_single_splat("SELECT COUNT(*) FROM users")
puts "Row count: #{row_count}" if row_count != ROW_COUNT
db.close
ensure
tempfile.close
tempfile.unlink
end
class SingleWriter
def initialize(db_path, row_count)
@row_count = row_count
@db = create_extralite_db(db_path)
@stmt = @db.prepare(SQL_INSERT)
end
def write
@row_count.times { @stmt.execute(USER) }
@stmt.close
@db.close
end
end
class ForkedSameDbWriter
def initialize(db_path, row_count)
@row_count = row_count
@db_path = db_path
@pids = []
setup_forks
end
def setup_forks
fork_count = Etc.nprocessors
split_row_count = @row_count / fork_count
fork_count.times do
@pids << fork do
db = create_extralite_db(@db_path)
stmt = db.prepare(SQL_INSERT)
Signal.trap("USR1") do
split_row_count.times { stmt.execute(USER) }
stmt.close
db.close
exit
end
sleep
end
end
sleep(1)
end
def write
@pids.each { |pid| Process.kill("USR1", pid) }
Process.waitall
end
end
class ForkedMultiDbWriter
def initialize(db_path, row_count)
@row_count = row_count
@complete_db_path = db_path
@pids = []
@db_paths = []
@db = create_extralite_db(db_path)
setup_forks
end
def setup_forks
fork_count = Etc.nprocessors
split_row_count = @row_count / fork_count
fork_count.times do |i|
db_path = "#{@complete_db_path}-#{i}"
@db_paths << db_path
@pids << fork do
db = create_extralite_db(db_path, initialize: true)
stmt = db.prepare(SQL_INSERT)
Signal.trap("USR1") do
split_row_count.times { stmt.execute(USER) }
stmt.close
db.close
exit
end
sleep
end
end
sleep(2)
end
def write
@pids.each { |pid| Process.kill("USR1", pid) }
Process.waitall
@db_paths.each do |db_path|
@db.execute("ATTACH DATABASE ? AS db", db_path)
@db.execute("INSERT INTO users SELECT * FROM db.users")
@db.execute("DETACH DATABASE db")
end
@db.close
end
end
LABEL_WIDTH = 25
def benchmark(label, label_width = 15)
print "#{label} ..."
label = label.ljust(label_width)
start = Process.clock_gettime(Process::CLOCK_MONOTONIC)
yield
finish = Process.clock_gettime(Process::CLOCK_MONOTONIC)
time_diff = sprintf("%.4f", finish - start).rjust(9)
print "\r#{label} #{time_diff} seconds\n"
end
puts "", "Benchmarking write performance", ""
with_db_path do |db_path|
single_writer = SingleWriter.new(db_path, ROW_COUNT)
benchmark("single writer", LABEL_WIDTH) { single_writer.write }
end
with_db_path do |db_path|
forked_same_db_writer = ForkedSameDbWriter.new(db_path, ROW_COUNT)
benchmark("forked writer - same DB", LABEL_WIDTH) { forked_same_db_writer.write }
end
with_db_path do |db_path|
forked_multi_db_writer = ForkedMultiDbWriter.new(db_path, ROW_COUNT)
benchmark("forked writer - multi DB", LABEL_WIDTH) { forked_multi_db_writer.write }
end