提交 143fc48a 编写于 作者: L Lin Jen-Shin

Add RequestStoreWrap to cache via RequestStore

I don't like the idea of `RequestStore` at all, because it's just a
global state which shouldn't be used at all. But we have a number of
places calling `ProtectedBranch.protected?` and `ProtectedTag.protected?`
in a loop for the same user, project, and ref whenever we're checking
against if the jobs for a given pipeline is accessible for a given user.
This means we're effectively making N queries for the same thing over
and over.

To properly fix this, we need to change how we check the permission,
and that could be a huge work. To solve this quickly, adding a cache
layer for the given request would be quite simple to do.

We're already doing this in Commit#author, and this is extending that
idea and make it generalized.
上级 05329d4a
module Gitlab
module Cache
# This module provides a simple way to cache values in RequestStore,
# and the cache key would be based on the class name, method name,
# customized instance level values, and arguments.
#
# A simple example:
#
# class UserAccess
# extend Gitlab::Cache::RequestStoreWrap
#
# request_store_wrap_key do
# [user.id, project.id]
# end
#
# request_store_wrap def can_push_to_branch?(ref)
# # ...
# end
# end
#
# This way, the result of `can_push_to_branch?` would be cached in
# `RequestStore.store` based on the cache key.
module RequestStoreWrap
def self.extended(klass)
return if klass < self
extension = Module.new
klass.const_set(:RequestStoreWrapExtension, extension)
klass.prepend(extension)
end
def request_store_wrap_key(&block)
if block_given?
@request_store_wrap_key = block
else
@request_store_wrap_key
end
end
def request_store_wrap(method_name)
const_get(:RequestStoreWrapExtension)
.send(:define_method, method_name) do |*args|
return super(*args) unless RequestStore.active?
klass = self.class
key = [klass.name,
method_name,
*instance_exec(&klass.request_store_wrap_key),
*args].join(':')
if RequestStore.store.key?(key)
RequestStore.store[key]
else
RequestStore.store[key] = super(*args)
end
end
end
end
end
end
module Gitlab
class UserAccess
extend Gitlab::Cache::RequestStoreWrap
request_store_wrap_key do
[user&.id, project&.id]
end
attr_reader :user, :project
def initialize(user, project: nil)
......@@ -28,7 +34,7 @@ module Gitlab
true
end
def can_create_tag?(ref)
request_store_wrap def can_create_tag?(ref)
return false unless can_access_git?
if ProtectedTag.protected?(project, ref)
......@@ -38,7 +44,7 @@ module Gitlab
end
end
def can_delete_branch?(ref)
request_store_wrap def can_delete_branch?(ref)
return false unless can_access_git?
if ProtectedBranch.protected?(project, ref)
......@@ -48,7 +54,7 @@ module Gitlab
end
end
def can_push_to_branch?(ref)
request_store_wrap def can_push_to_branch?(ref)
return false unless can_access_git?
if ProtectedBranch.protected?(project, ref)
......@@ -60,7 +66,7 @@ module Gitlab
end
end
def can_merge_to_branch?(ref)
request_store_wrap def can_merge_to_branch?(ref)
return false unless can_access_git?
if ProtectedBranch.protected?(project, ref)
......
require 'spec_helper'
describe Gitlab::Cache::RequestStoreWrap, :request_store do
let(:klass) do
Class.new do
extend Gitlab::Cache::RequestStoreWrap
attr_accessor :id, :name, :result
def self.name
'ExpensiveAlgorithm'
end
def initialize(id, name, result)
self.id = id
self.name = name
self.result = result
end
request_store_wrap_key do
[id, name]
end
request_store_wrap def compute(arg)
result << arg
end
request_store_wrap def repute(arg)
result << arg
end
end
end
let(:algorithm) { klass.new('id', 'name', []) }
context 'when RequestStore is active' do
it 'does not compute twice for the same argument' do
result = algorithm.compute(true)
expect(result).to eq([true])
expect(algorithm.compute(true)).to eq(result)
expect(algorithm.result).to eq(result)
end
it 'computes twice for the different argument' do
algorithm.compute(true)
result = algorithm.compute(false)
expect(result).to eq([true, false])
expect(algorithm.result).to eq(result)
end
it 'computes twice for the different keys, id' do
algorithm.compute(true)
algorithm.id = 'ad'
result = algorithm.compute(true)
expect(result).to eq([true, true])
expect(algorithm.result).to eq(result)
end
it 'computes twice for the different keys, name' do
algorithm.compute(true)
algorithm.name = 'same'
result = algorithm.compute(true)
expect(result).to eq([true, true])
expect(algorithm.result).to eq(result)
end
it 'computes twice for the different class name' do
algorithm.compute(true)
allow(klass).to receive(:name).and_return('CheapAlgo')
result = algorithm.compute(true)
expect(result).to eq([true, true])
expect(algorithm.result).to eq(result)
end
it 'computes twice for the different method' do
algorithm.compute(true)
result = algorithm.repute(true)
expect(result).to eq([true, true])
expect(algorithm.result).to eq(result)
end
it 'computes twice if RequestStore starts over' do
algorithm.compute(true)
RequestStore.end!
RequestStore.clear!
RequestStore.begin!
result = algorithm.compute(true)
expect(result).to eq([true, true])
expect(algorithm.result).to eq(result)
end
end
context 'when RequestStore is inactive' do
before do
RequestStore.end!
end
it 'computes twice even if everything is the same' do
algorithm.compute(true)
result = algorithm.compute(true)
expect(result).to eq([true, true])
expect(algorithm.result).to eq(result)
end
end
end
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册