Update links to upstream migration helpers, remove unused methods (#28781)

This commit is contained in:
Matt Jankowski 2024-01-18 07:29:54 -05:00 committed by GitHub
parent aaa6f2e930
commit 81e4e65610
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -8,15 +8,15 @@
# shorten temporary column names. # shorten temporary column names.
# Documentation on using these functions (and why one might do so): # Documentation on using these functions (and why one might do so):
# https://gitlab.com/gitlab-org/gitlab-ce/blob/master/doc/development/what_requires_downtime.md # https://gitlab.com/gitlab-org/gitlab-foss/-/blob/master/doc/development/database/avoiding_downtime_in_migrations.md
# The file itself: # The original file (since updated):
# https://gitlab.com/gitlab-org/gitlab-ce/blob/master/lib/gitlab/database/migration_helpers.rb # https://gitlab.com/gitlab-org/gitlab-foss/-/blob/master/lib/gitlab/database/migration_helpers.rb
# It is licensed as follows: # It is licensed as follows:
# Copyright (c) 2011-2017 GitLab B.V. # Copyright (c) 2011-present GitLab B.V.
#
# Permission is hereby granted, free of charge, to any person obtaining a copy # Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal # of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights # in the Software without restriction, including without limitation the rights
@ -24,16 +24,16 @@
# copies of the Software, and to permit persons to whom the Software is # copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions: # furnished to do so, subject to the following conditions:
# The above copyright notice and this permission notice shall be included in # The above copyright notice and this permission notice shall be included in all
# all copies or substantial portions of the Software. # copies or substantial portions of the Software.
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# THE SOFTWARE. # SOFTWARE.
# This is bad form, but there are enough differences that it's impractical to do # This is bad form, but there are enough differences that it's impractical to do
# otherwise: # otherwise:
@ -77,37 +77,12 @@ module Mastodon
end end
end end
BACKGROUND_MIGRATION_BATCH_SIZE = 1000 # Number of rows to process per job
BACKGROUND_MIGRATION_JOB_BUFFER_SIZE = 1000 # Number of jobs to bulk queue at a time
# Gets an estimated number of rows for a table # Gets an estimated number of rows for a table
def estimate_rows_in_table(table_name) def estimate_rows_in_table(table_name)
exec_query('SELECT reltuples FROM pg_class WHERE relname = ' + exec_query('SELECT reltuples FROM pg_class WHERE relname = ' +
"'#{table_name}'").to_a.first['reltuples'] "'#{table_name}'").to_a.first['reltuples']
end end
# Adds `created_at` and `updated_at` columns with timezone information.
#
# This method is an improved version of Rails' built-in method `add_timestamps`.
#
# Available options are:
# default - The default value for the column.
# null - When set to `true` the column will allow NULL values.
# The default is to not allow NULL values.
def add_timestamps_with_timezone(table_name, **options)
options[:null] = false if options[:null].nil?
[:created_at, :updated_at].each do |column_name|
if options[:default] && transaction_open?
raise '`add_timestamps_with_timezone` with default value cannot be run inside a transaction. ' \
'You can disable transactions by calling `disable_ddl_transaction!` ' \
'in the body of your migration class'
end
add_column(table_name, column_name, :datetime_with_timezone, **options)
end
end
# Creates a new index, concurrently when supported # Creates a new index, concurrently when supported
# #
# On PostgreSQL this method creates an index concurrently, on MySQL this # On PostgreSQL this method creates an index concurrently, on MySQL this
@ -746,39 +721,6 @@ module Mastodon
rename_index table_name, "#{index_name}_new", index_name rename_index table_name, "#{index_name}_new", index_name
end end
# This will replace the first occurrence of a string in a column with
# the replacement
# On postgresql we can use `regexp_replace` for that.
# On mysql we find the location of the pattern, and overwrite it
# with the replacement
def replace_sql(column, pattern, replacement)
quoted_pattern = Arel::Nodes::Quoted.new(pattern.to_s)
quoted_replacement = Arel::Nodes::Quoted.new(replacement.to_s)
replace = Arel::Nodes::NamedFunction
.new("regexp_replace", [column, quoted_pattern, quoted_replacement])
Arel::Nodes::SqlLiteral.new(replace.to_sql)
end
def remove_foreign_key_without_error(*args)
remove_foreign_key(*args)
rescue ArgumentError
end
def sidekiq_queue_migrate(queue_from, to:)
while sidekiq_queue_length(queue_from) > 0
Sidekiq.redis do |conn|
conn.rpoplpush "queue:#{queue_from}", "queue:#{to}"
end
end
end
def sidekiq_queue_length(queue_name)
Sidekiq.redis do |conn|
conn.llen("queue:#{queue_name}")
end
end
def check_trigger_permissions!(table) def check_trigger_permissions!(table)
unless Grant.create_and_execute_trigger?(table) unless Grant.create_and_execute_trigger?(table)
dbname = ActiveRecord::Base.configurations[Rails.env]['database'] dbname = ActiveRecord::Base.configurations[Rails.env]['database']
@ -799,91 +741,6 @@ into similar problems in the future (e.g. when new tables are created).
end end
end end
# Bulk queues background migration jobs for an entire table, batched by ID range.
# "Bulk" meaning many jobs will be pushed at a time for efficiency.
# If you need a delay interval per job, then use `queue_background_migration_jobs_by_range_at_intervals`.
#
# model_class - The table being iterated over
# job_class_name - The background migration job class as a string
# batch_size - The maximum number of rows per job
#
# Example:
#
# class Route < ActiveRecord::Base
# include EachBatch
# self.table_name = 'routes'
# end
#
# bulk_queue_background_migration_jobs_by_range(Route, 'ProcessRoutes')
#
# Where the model_class includes EachBatch, and the background migration exists:
#
# class Gitlab::BackgroundMigration::ProcessRoutes
# def perform(start_id, end_id)
# # do something
# end
# end
def bulk_queue_background_migration_jobs_by_range(model_class, job_class_name, batch_size: BACKGROUND_MIGRATION_BATCH_SIZE)
raise "#{model_class} does not have an ID to use for batch ranges" unless model_class.column_names.include?('id')
jobs = []
model_class.each_batch(of: batch_size) do |relation|
start_id, end_id = relation.pluck('MIN(id), MAX(id)').first
if jobs.length >= BACKGROUND_MIGRATION_JOB_BUFFER_SIZE
# Note: This code path generally only helps with many millions of rows
# We push multiple jobs at a time to reduce the time spent in
# Sidekiq/Redis operations. We're using this buffer based approach so we
# don't need to run additional queries for every range.
BackgroundMigrationWorker.perform_bulk(jobs)
jobs.clear
end
jobs << [job_class_name, [start_id, end_id]]
end
BackgroundMigrationWorker.perform_bulk(jobs) unless jobs.empty?
end
# Queues background migration jobs for an entire table, batched by ID range.
# Each job is scheduled with a `delay_interval` in between.
# If you use a small interval, then some jobs may run at the same time.
#
# model_class - The table being iterated over
# job_class_name - The background migration job class as a string
# delay_interval - The duration between each job's scheduled time (must respond to `to_f`)
# batch_size - The maximum number of rows per job
#
# Example:
#
# class Route < ActiveRecord::Base
# include EachBatch
# self.table_name = 'routes'
# end
#
# queue_background_migration_jobs_by_range_at_intervals(Route, 'ProcessRoutes', 1.minute)
#
# Where the model_class includes EachBatch, and the background migration exists:
#
# class Gitlab::BackgroundMigration::ProcessRoutes
# def perform(start_id, end_id)
# # do something
# end
# end
def queue_background_migration_jobs_by_range_at_intervals(model_class, job_class_name, delay_interval, batch_size: BACKGROUND_MIGRATION_BATCH_SIZE)
raise "#{model_class} does not have an ID to use for batch ranges" unless model_class.column_names.include?('id')
model_class.each_batch(of: batch_size) do |relation, index|
start_id, end_id = relation.pluck('MIN(id), MAX(id)').first
# `BackgroundMigrationWorker.bulk_perform_in` schedules all jobs for
# the same time, which is not helpful in most cases where we wish to
# spread the work over time.
BackgroundMigrationWorker.perform_in(delay_interval * index, job_class_name, [start_id, end_id])
end
end
private private
# https://github.com/rails/rails/blob/v5.2.0/activerecord/lib/active_record/connection_adapters/postgresql/schema_statements.rb#L678-L684 # https://github.com/rails/rails/blob/v5.2.0/activerecord/lib/active_record/connection_adapters/postgresql/schema_statements.rb#L678-L684