Commit d9efc7d9 authored by Timothy Andrew's avatar Timothy Andrew

Merge remote-tracking branch 'origin/master' into ee-581-backport-changes

parents 8c101fc3 7f853e22
......@@ -117,6 +117,9 @@ v 8.11.0 (unreleased)
- Speed up todos queries by limiting the projects set we join with
- Ensure file editing in UI does not overwrite commited changes without warning user
v 8.10.6 (unreleased)
- Fix import/export configuration missing some included attributes
v 8.10.5
- Add a data migration to fix some missing timestamps in the members table. !5670
- Revert the "Defend against 'Host' header injection" change in the source NGINX templates. !5706
......
class Import::GitlabProjectsController < Import::BaseController
before_action :verify_gitlab_project_import_enabled
before_action :authenticate_admin!
def new
@namespace_id = project_params[:namespace_id]
......@@ -47,4 +48,8 @@ def project_params
:path, :namespace_id, :file
)
end
def authenticate_admin!
render_404 unless current_user.is_admin?
end
end
......@@ -77,7 +77,7 @@
= link_to "#", class: 'btn js-toggle-button import_git' do
= icon('git', text: 'Repo by URL')
%div{ class: 'import_gitlab_project' }
- if gitlab_project_import_enabled?
- if gitlab_project_import_enabled? && current_user.is_admin?
= link_to new_import_gitlab_project_path, class: 'btn btn_import_gitlab_project project-submit' do
= icon('gitlab', text: 'GitLab export')
......
......@@ -30,7 +30,11 @@
- [Rake tasks](rake_tasks.md) for development
- [Shell commands](shell_commands.md) in the GitLab codebase
- [Sidekiq debugging](sidekiq_debugging.md)
## Databases
- [What requires downtime?](what_requires_downtime.md)
- [Adding database indexes](adding_database_indexes.md)
## Compliance
......
# Adding Database Indexes
Indexes can be used to speed up database queries, but when should you add a new
index? Traditionally the answer to this question has been to add an index for
every column used for filtering or joining data. For example, consider the
following query:
```sql
SELECT *
FROM projects
WHERE user_id = 2;
```
Here we are filtering by the `user_id` column and as such a developer may decide
to index this column.
While in certain cases indexing columns using the above approach may make sense
it can actually have a negative impact. Whenever you write data to a table any
existing indexes need to be updated. The more indexes there are the slower this
can potentially become. Indexes can also take up quite some disk space depending
on the amount of data indexed and the index type. For example, PostgreSQL offers
"GIN" indexes which can be used to index certain data types that can not be
indexed by regular btree indexes. These indexes however generally take up more
data and are slower to update compared to btree indexes.
Because of all this one should not blindly add a new index for every column used
to filter data by. Instead one should ask themselves the following questions:
1. Can I write my query in such a way that it re-uses as many existing indexes
as possible?
2. Is the data going to be large enough that using an index will actually be
faster than just iterating over the rows in the table?
3. Is the overhead of maintaining the index worth the reduction in query
timings?
We'll explore every question in detail below.
## Re-using Queries
The first step is to make sure your query re-uses as many existing indexes as
possible. For example, consider the following query:
```sql
SELECT *
FROM todos
WHERE user_id = 123
AND state = 'open';
```
Now imagine we already have an index on the `user_id` column but not on the
`state` column. One may think this query will perform badly due to `state` being
unindexed. In reality the query may perform just fine given the index on
`user_id` can filter out enough rows.
The best way to determine if indexes are re-used is to run your query using
`EXPLAIN ANALYZE`. Depending on any extra tables that may be joined and
other columns being used for filtering you may find an extra index is not going
to make much (if any) difference. On the other hand you may determine that the
index _may_ make a difference.
In short:
1. Try to write your query in such a way that it re-uses as many existing
indexes as possible.
2. Run the query using `EXPLAIN ANALYZE` and study the output to find the most
ideal query.
## Data Size
A database may decide not to use an index despite it existing in case a regular
sequence scan (= simply iterating over all existing rows) is faster. This is
especially the case for small tables.
If a table is expected to grow in size and you expect your query has to filter
out a lot of rows you may want to consider adding an index. If the table size is
very small (e.g. only a handful of rows) or any existing indexes filter out
enough rows you may _not_ want to add a new index.
## Maintenance Overhead
Indexes have to be updated on every table write. In case of PostgreSQL _all_
existing indexes will be updated whenever data is written to a table. As a
result of this having many indexes on the same table will slow down writes.
Because of this one should ask themselves: is the reduction in query performance
worth the overhead of maintaining an extra index?
If adding an index reduces SELECT timings by 5 milliseconds but increases
INSERT/UPDATE/DELETE timings by 10 milliseconds then the index may not be worth
it. On the other hand, if SELECT timings are reduced but INSERT/UPDATE/DELETE
timings are not affected you may want to add the index after all.
## Finding Unused Indexes
To see which indexes are unused you can run the following query:
```sql
SELECT relname as table_name, indexrelname as index_name, idx_scan, idx_tup_read, idx_tup_fetch, pg_size_pretty(pg_relation_size(indexrelname::regclass))
FROM pg_stat_all_indexes
WHERE schemaname = 'public'
AND idx_scan = 0
AND idx_tup_read = 0
AND idx_tup_fetch = 0
ORDER BY pg_relation_size(indexrelname::regclass) desc;
```
This query outputs a list containing all indexes that are never used and sorts
them by indexes sizes in descending order. This query can be useful to
determine if any previously indexes are useful after all. More information on
the meaning of the various columns can be found at
<https://www.postgresql.org/docs/current/static/monitoring-stats.html>.
Because the output of this query relies on the actual usage of your database it
may be affected by factors such as (but not limited to):
* Certain queries never being executed, thus not being able to use certain
indexes.
* Certain tables having little data, resulting in PostgreSQL using sequence
scans instead of index scans.
In other words, this data is only reliable for a frequently used database with
plenty of data and with as many GitLab features enabled (and being used) as
possible.
......@@ -15,11 +15,14 @@ repository and maintained by GitLab UX designers.
## Navigation
GitLab's layout contains 2 sections: the left sidebar and the content. The left sidebar contains a static navigation menu.
This menu will be visible regardless of what page you visit. The left sidebar also contains the GitLab logo
and the current user's profile picture. The content section contains a header and the content itself.
The header describes the current GitLab page and what navigation is
available to user in this area. Depending on the area (project, group, profile setting) the header name and navigation may change. For example when user visits one of the
project pages the header will contain a project name and navigation for that project. When the user visits a group page it will contain a group name and navigation related to this group.
This menu will be visible regardless of what page you visit.
The content section contains a header and the content itself. The header describes the current GitLab page and what navigation is
available to the user in this area. Depending on the area (project, group, profile setting) the header name and navigation may change. For example, when the user visits one of the
project pages the header will contain the project's name and navigation for that project. When the user visits a group page it will contain the group's name and navigation related to this group.
You can see a visual representation of the navigation in GitLab in the GitLab Product Map, which is located in the [Design Repository](gitlab-map-graffle)
along with [PDF](gitlab-map-pdf) and [PNG](gitlab-map-png) exports.
### Adding new tab to header navigation
......@@ -99,3 +102,6 @@ Do not use both green and blue button in one form.
display counts in the UI.
[number_with_delimiter]: http://api.rubyonrails.org/classes/ActionView/Helpers/NumberHelper.html#method-i-number_with_delimiter
[gitlab-map-graffle]: https://gitlab.com/gitlab-org/gitlab-design/blob/master/production/resources/gitlab-map.graffle
[gitlab-map-pdf]: https://gitlab.com/gitlab-org/gitlab-design/raw/master/production/resources/gitlab-map.pdf
[gitlab-map-png]: https://gitlab.com/gitlab-org/gitlab-design/raw/master/production/resources/gitlab-map.png
\ No newline at end of file
......@@ -7,8 +7,7 @@
> than that of the exporter.
> - For existing installations, the project import option has to be enabled in
> application settings (`/admin/application_settings`) under 'Import sources'.
> Ask your administrator if you don't see the **GitLab export** button when
> creating a new project.
> You will have to be an administrator to enable and use the import functionality.
> - You can find some useful raketasks if you are an administrator in the
> [import_export](../../../administration/raketasks/project_import_export.md)
> raketask.
......
......@@ -9,7 +9,7 @@ Background:
@javascript
Scenario: I should see New Projects page
Then I see "New Project" page
Then I see all possible import optios
Then I see all possible import options
@javascript
Scenario: I should see instructions on how to import from Git URL
......
......@@ -106,6 +106,16 @@ Feature: Project Merge Requests
And I sort the list by "Least popular"
Then The list should be sorted by "Least popular"
@javascript
Scenario: I comment on a merge request diff
Given project "Shop" have "Bug NS-05" open merge request with diffs inside
And I visit merge request page "Bug NS-05"
And I click on the Changes tab
And I leave a comment like "Line is wrong" on diff
And I switch to the merge request's comments tab
Then I should see a discussion has started on diff
And I should see a badge of "1" next to the discussion link
@javascript
Scenario: I see a new comment on merge request diff from another user in the discussion tab
Given project "Shop" have "Bug NS-05" open merge request with diffs inside
......
......@@ -14,14 +14,13 @@ class Spinach::Features::NewProject < Spinach::FeatureSteps
expect(page).to have_content('Project name')
end
step 'I see all possible import optios' do
step 'I see all possible import options' do
expect(page).to have_link('GitHub')
expect(page).to have_link('Bitbucket')
expect(page).to have_link('GitLab.com')
expect(page).to have_link('Gitorious.org')
expect(page).to have_link('Google Code')
expect(page).to have_link('Repo by URL')
expect(page).to have_link('GitLab export')
end
step 'I click on "Import project from GitHub"' do
......
......@@ -57,19 +57,16 @@ def handle_model_object(current_key, model_object, json_config_hash)
# +value+ existing model to be included in the hash
# +json_config_hash+ the original hash containing the root model
def create_model_value(current_key, value, json_config_hash)
parsed_hash = { include: value }
parse_hash(value, parsed_hash)
json_config_hash[current_key] = parsed_hash
json_config_hash[current_key] = parse_hash(value) || { include: value }
end
# Calls attributes finder to parse the hash and add any attributes to it
#
# +value+ existing model to be included in the hash
# +parsed_hash+ the original hash
def parse_hash(value, parsed_hash)
def parse_hash(value)
@attributes_finder.parse(value) do |hash|
parsed_hash = { include: hash_or_merge(value, hash) }
{ include: hash_or_merge(value, hash) }
end
end
......
require 'spec_helper'
feature 'Diff notes', js: true, feature: true do
include WaitForAjax
before do
login_as :admin
@merge_request = create(:merge_request)
@project = @merge_request.source_project
end
context 'merge request diffs' do
let(:comment_button_class) { '.add-diff-note' }
let(:notes_holder_input_class) { 'js-temp-notes-holder' }
let(:notes_holder_input_xpath) { './following-sibling::*[contains(concat(" ", @class, " "), " notes_holder ")]' }
let(:test_note_comment) { 'this is a test note!' }
context 'when hovering over the parallel view diff file' do
before(:each) do
visit diffs_namespace_project_merge_request_path(@project.namespace, @project, @merge_request)
click_link 'Side-by-side'
end
context 'with an old line on the left and no line on the right' do
it 'should allow commenting on the left side' do
should_allow_commenting(find('[id="6eb14e00385d2fb284765eb1cd8d420d33d63fc9_23_22"]').find(:xpath, '..'), 'left')
end
it 'should not allow commenting on the right side' do
should_not_allow_commenting(find('[id="6eb14e00385d2fb284765eb1cd8d420d33d63fc9_23_22"]').find(:xpath, '..'), 'right')
end
end
context 'with no line on the left and a new line on the right' do
it 'should not allow commenting on the left side' do
should_not_allow_commenting(find('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_15_15"]').find(:xpath, '..'), 'left')
end
it 'should allow commenting on the right side' do
should_allow_commenting(find('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_15_15"]').find(:xpath, '..'), 'right')
end
end
context 'with an old line on the left and a new line on the right' do
it 'should allow commenting on the left side' do
should_allow_commenting(find('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_9_9"]').find(:xpath, '..'), 'left')
end
it 'should allow commenting on the right side' do
should_allow_commenting(find('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_9_9"]').find(:xpath, '..'), 'right')
end
end
context 'with an unchanged line on the left and an unchanged line on the right' do
it 'should allow commenting on the left side' do
should_allow_commenting(first('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_7_7"]').find(:xpath, '..'), 'left')
end
it 'should allow commenting on the right side' do
should_allow_commenting(first('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_7_7"]').find(:xpath, '..'), 'right')
end
end
context 'with a match line' do
it 'should not allow commenting on the left side' do
should_not_allow_commenting(first('.match').find(:xpath, '..'), 'left')
end
it 'should not allow commenting on the right side' do
should_not_allow_commenting(first('.match').find(:xpath, '..'), 'right')
end
end
end
context 'when hovering over the inline view diff file' do
before do
visit diffs_namespace_project_merge_request_path(@project.namespace, @project, @merge_request)
click_link 'Inline'
end
context 'with a new line' do
it 'should allow commenting' do
should_allow_commenting(find('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_10_9"]'))
end
end
context 'with an old line' do
it 'should allow commenting' do
should_allow_commenting(find('[id="6eb14e00385d2fb284765eb1cd8d420d33d63fc9_22_22"]'))
end
end
context 'with an unchanged line' do
it 'should allow commenting' do
should_allow_commenting(find('[id="2f6fcd96b88b36ce98c38da085c795a27d92a3dd_7_7"]'))
end
end
context 'with a match line' do
it 'should not allow commenting' do
should_not_allow_commenting(first('.match'))
end
end
end
def should_allow_commenting(line_holder, diff_side = nil)
line = get_line_components(line_holder, diff_side)
line[:content].hover
expect(line[:num]).to have_css comment_button_class
comment_on_line(line_holder, line)
wait_for_ajax
assert_comment_persistence(line_holder)
end
def should_not_allow_commenting(line_holder, diff_side = nil)
line = get_line_components(line_holder, diff_side)
line[:content].hover
expect(line[:num]).not_to have_css comment_button_class
end
def get_line_components(line_holder, diff_side = nil)
if diff_side.nil?
get_inline_line_components(line_holder)
else
get_parallel_line_components(line_holder, diff_side)
end
end
def get_inline_line_components(line_holder)
{ content: line_holder.first('.line_content'), num: line_holder.first('.diff-line-num') }
end
def get_parallel_line_components(line_holder, diff_side = nil)
side_index = diff_side == 'left' ? 0 : 1
{ content: line_holder.all('.line_content')[side_index], num: line_holder.all('.diff-line-num')[side_index] }
end
def comment_on_line(line_holder, line)
line[:num].find(comment_button_class).trigger 'click'
expect(line_holder).to have_xpath notes_holder_input_xpath
notes_holder_input = line_holder.find(:xpath, notes_holder_input_xpath)
expect(notes_holder_input[:class]).to include(notes_holder_input_class)
notes_holder_input.fill_in 'note[note]', with: test_note_comment
click_button 'Comment'
end
def assert_comment_persistence(line_holder)
expect(line_holder).to have_xpath notes_holder_input_xpath
notes_holder_saved = line_holder.find(:xpath, notes_holder_input_xpath)
expect(notes_holder_saved[:class]).not_to include(notes_holder_input_class)
expect(notes_holder_saved).to have_content test_note_comment
end
end
end
......@@ -3,8 +3,9 @@
feature 'project import', feature: true, js: true do
include Select2Helper
let(:user) { create(:admin) }
let!(:namespace) { create(:namespace, name: "asd", owner: user) }
let(:admin) { create(:admin) }
let(:normal_user) { create(:user) }
let!(:namespace) { create(:namespace, name: "asd", owner: admin) }
let(:file) { File.join(Rails.root, 'spec', 'features', 'projects', 'import_export', 'test_project_export.tar.gz') }
let(:export_path) { "#{Dir::tmpdir}/import_file_spec" }
let(:project) { Project.last }
......@@ -12,66 +13,87 @@
background do
allow_any_instance_of(Gitlab::ImportExport).to receive(:storage_path).and_return(export_path)
login_as(user)
end
after(:each) do
FileUtils.rm_rf(export_path, secure: true)
end
scenario 'user imports an exported project successfully' do
expect(Project.all.count).to be_zero
context 'admin user' do
before do
login_as(admin)
end
visit new_project_path
scenario 'user imports an exported project successfully' do
expect(Project.all.count).to be_zero
select2('2', from: '#project_namespace_id')
fill_in :project_path, with: 'test-project-path', visible: true
click_link 'GitLab export'
visit new_project_path
expect(page).to have_content('GitLab project export')
expect(URI.parse(current_url).query).to eq('namespace_id=2&path=test-project-path')
select2('2', from: '#project_namespace_id')
fill_in :project_path, with: 'test-project-path', visible: true
click_link 'GitLab export'
attach_file('file', file)
expect(page).to have_content('GitLab project export')
expect(URI.parse(current_url).query).to eq('namespace_id=2&path=test-project-path')
click_on 'Import project' # import starts
attach_file('file', file)
expect(project).not_to be_nil
expect(project.issues).not_to be_empty
expect(project.merge_requests).not_to be_empty
expect(project_hook).to exist
expect(wiki_exists?).to be true
expect(project.import_status).to eq('finished')
end
click_on 'Import project' # import starts
expect(project).not_to be_nil
expect(project.issues).not_to be_empty
expect(project.merge_requests).not_to be_empty
expect(project_hook).to exist
expect(wiki_exists?).to be true
expect(project.import_status).to eq('finished')
end
scenario 'invalid project' do
project = create(:project, namespace_id: 2)
scenario 'invalid project' do
project = create(:project, namespace_id: 2)
visit new_project_path
visit new_project_path
select2('2', from: '#project_namespace_id')
fill_in :project_path, with: project.name, visible: true
click_link 'GitLab export'
select2('2', from: '#project_namespace_id')
fill_in :project_path, with: project.name, visible: true
click_link 'GitLab export'
attach_file('file', file)
click_on 'Import project'
attach_file('file', file)
click_on 'Import project'
page.within('.flash-container') do
expect(page).to have_content('Project could not be imported')
page.within('.flash-container') do
expect(page).to have_content('Project could not be imported')
end
end
scenario 'project with no name' do
create(:project, namespace_id: 2)
visit new_project_path
select2('2', from: '#project_namespace_id')
# click on disabled element
find(:link, 'GitLab export').trigger('click')
page.within('.flash-container') do
expect(page).to have_content('Please enter path and name')
end
end
end
scenario 'project with no name' do
create(:project, namespace_id: 2)
context 'normal user' do
before do
login_as(normal_user)
end
visit new_project_path
scenario 'non-admin user is not allowed to import a project' do
expect(Project.all.count).to be_zero
select2('2', from: '#project_namespace_id')
visit new_project_path
# click on disabled element
find(:link, 'GitLab export').trigger('click')
fill_in :project_path, with: 'test-project-path', visible: true
page.within('.flash-container') do
expect(page).to have_content('Please enter path and name')
expect(page).not_to have_content('GitLab export')
end
end
......
......@@ -12,7 +12,8 @@
except: [:iid],
include: [:merge_request_diff, :merge_request_test]
} },
{ commit_statuses: { include: :commit } }]
{ commit_statuses: { include: :commit } },
{ project_members: { include: { user: { only: [:email] } } } }]
}
end
......
......@@ -7,6 +7,8 @@ project_tree:
- :merge_request_test
- commit_statuses:
- :commit
- project_members:
- :user
included_attributes:
project:
......@@ -14,6 +16,8 @@ included_attributes:
- :path
merge_requests:
- :id
user:
- :email
excluded_attributes:
merge_requests:
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment