Improve restoring of access entities' dependencies #2 by vitlibar · Pull Request #69563 · ClickHouse/ClickHouse

vitlibar · 2024-09-12T15:48:42Z

Changelog category:

Improvement

Changelog entry:

Follow-up to #69346
Point 4 described there will work now as well:

For example if before making a backup a role was granted to a user:

CREATE USER u1;
CREATE ROLE r1;
GRANT r1 TO u1;

then during RESTORE the following cases are possible:

User u1 and role r1 both don't exist. Then RESTORE will recreate them both and that grant too.
User u1 and role r1 both exist. Then RESTORE (with default settings) will skip them both and that grant too.
User u1 doesn't exist and role r1 exist. Then RESTORE (with default settings) will recreate user u1, and that grant too.
User u1 exists and role r1 doesn't exist. Then RESTORE (with default settings) will recreate role r1, and (after this PR) that grant too.

robot-clickhouse · 2024-09-12T15:51:35Z

This is an automated comment for commit 8a7a411 with description of existing statuses. It's updated for the latest CI running

✅ Click here to open a full report in a separate page

Successful checks

Check name	Description	Status
AST fuzzer	Runs randomly generated queries to catch program errors. The build type is optionally given in parenthesis. If it fails, ask a maintainer for help	✅ success
Builds	There's no description for the check yet, please add it to tests/ci/ci_config.py:CHECK_DESCRIPTIONS	✅ success
ClickBench	Runs [ClickBench](https://github.com/ClickHouse/ClickBench/) with instant-attach table	✅ success
Compatibility check	Checks that clickhouse binary runs on distributions with old libc versions. If it fails, ask a maintainer for help	✅ success
Docker keeper image	The check to build and optionally push the mentioned image to docker hub	✅ success
Docker server image	The check to build and optionally push the mentioned image to docker hub	✅ success
Docs check	Builds and tests the documentation	✅ success
Fast test	Normally this is the first check that is ran for a PR. It builds ClickHouse and runs most of stateless functional tests, omitting some. If it fails, further checks are not started until it is fixed. Look at the report to see which tests fail, then reproduce the failure locally as described here	✅ success
Flaky tests	Checks if new added or modified tests are flaky by running them repeatedly, in parallel, with more randomization. Functional tests are run 100 times with address sanitizer, and additional randomization of thread scheduling. Integration tests are run up to 10 times. If at least once a new test has failed, or was too long, this check will be red. We don't allow flaky tests, read the doc	✅ success
Install packages	Checks that the built packages are installable in a clear environment	✅ success
Integration tests	The integration tests report. In parenthesis the package type is given, and in square brackets are the optional part/total tests	✅ success
Performance Comparison	Measure changes in query performance. The performance test report is described in detail here. In square brackets are the optional part/total tests	✅ success
Stateful tests	Runs stateful functional tests for ClickHouse binaries built in various configurations -- release, debug, with sanitizers, etc	✅ success
Stateless tests	Runs stateless functional tests for ClickHouse binaries built in various configurations -- release, debug, with sanitizers, etc	✅ success
Stress test	Runs stateless functional tests concurrently from several clients to detect concurrency-related errors	✅ success
Style check	Runs a set of checks to keep the code style clean. If some of tests failed, see the related log from the report	✅ success
Unit tests	Runs the unit tests for different release types	✅ success
Upgrade check	Runs stress tests on server version from last release and then tries to upgrade it to the version from the PR. It checks if the new server can successfully startup without any errors, crashes or sanitizer asserts	✅ success

qoega · 2024-09-16T08:39:25Z

CI all green

vitlibar · 2024-09-16T09:35:52Z

I'd like to merge #69346 first.

…ties.

…_access_dependencies = true".

…updating dependents of restored access entities.

use "access-<UUID>.txt" instead of "access01.txt", "access02.txt", ...

…n the storage.

… setting "write_access_entities_dependents".

…ependents too.

…p_unresolved_access_dependencies".

vitlibar · 2024-09-23T11:07:26Z

@pufit Please review

pufit

Overall, looks good!

pufit · 2024-09-26T05:51:16Z

src/Access/IAccessStorage.h

-    std::vector<UUID> findAll() const { return findAll(EntityClassT::TYPE); }
+    /// Returns the identifiers of all the entities in the storage.
+    template <typename EntityClassT = IAccessEntity>
+    std::vector<UUID> findAll() const;


do we use it somewhere?

yes, see https://github.com/ClickHouse/ClickHouse/pull/69563/files#diff-5b07fbe0e42dd2e297b2aaa50ebb5545f132ab5c3884d38b94c4d7cb06f7151dR913

When we're making a backup:

first it collects all the access entities with BackupEntriesCollector::getAllAccessEntities(); then

IAccessStorage::backup() finds UUIDs of all entities of the required type (users/roles/etc); and finally

makeBackupEntryForAccessEntities() produces backup entities for the backup.

pufit · 2024-09-26T05:57:58Z

src/Access/AccessBackup.cpp

+            if (!entities_ids_set.contains(id) && possible_dependent->hasDependencies(entities_ids_set))
+            {
+                auto dependent = possible_dependent->clone();
+                dependent->clearAllExceptDependencies();


What does it do exactly? Why do we need to remove everything here?

This is to store dependent access entities. For example, if we store roles:

BACKUP TABLE system.roles ...

then after this PR we'll have additional DEPENDENTS section in the backup with information about which users use those roles. This information must not contain the credentials of these users.

For example, if a role was granted to a user and if we backup system.roles then file /data/system/roles/access-<UUID>.txt inside the backup will look like this:

fb4c9f55-3c15-32b3-4c5b-ea6de397c48e ROLE role1 ATTACH ROLE role1 SETTINGS custom_x = 'x'; DEPENDENTS 7647694e-1f79-5d14-dd2c-29d976450ba1 USER user1 ATTACH USER user1 DEFAULT ROLE ID('fb4c9f55-3c15-32b3-4c5b-ea6de397c48e'); ATTACH GRANT ID('fb4c9f55-3c15-32b3-4c5b-ea6de397c48e') TO user1;

So this information is to restore role role1 and grant it to existing user user1, and not to restore user1 - so we don't have to store the credentials of user1 here.

pufit · 2024-09-26T06:03:24Z

src/Backups/BackupCoordinationReplicatedAccess.h

    struct FilePathsAndHost
    {
-        std::unordered_set<String> file_paths;
+        std::set<String> file_paths;


why does it need to be sorted?

Because in BackupCoordinationReplicatedAccess::getFilePaths() we get the first file_path in this list and we want it to be the same on each host. This list of file paths contains the same elements on each host and if the list is sorted then we can be sure that the first element will be also the same on each host.

pufit · 2024-09-26T06:08:11Z

src/Access/GrantedRoles.cpp

+
+    if (found)
+    {
+        for (const auto & role_id : src.roles_with_admin_option)


pufit · 2024-09-26T06:11:41Z

src/Access/GrantedRoles.cpp

+
+    if (found)
+    {
+        for (auto it = roles_with_admin_option.begin(); it != roles_with_admin_option.end();)


…encies #2

vitlibar · 2024-09-28T19:56:51Z

I'm not sure about backport though - this PR has quite a bunch of changes, it adds new behavior and it even changes the format of some files stored inside a backup a little (in a backward-compatible manner).

alexey-milovidov · 2024-09-28T21:03:47Z

Yes, let's close backports.

Backport #69563 to 24.9: Improve restoring of access entities' dependencies #2

vitlibar mentioned this pull request Sep 12, 2024

Improve restoring of access entities' dependencies #69346

Merged

robot-ch-test-poll added the pr-improvement Pull request with some product improvements label Sep 12, 2024

vitlibar added the do not test disable testing on pull request label Sep 12, 2024

vitlibar force-pushed the restore-access-dependencies-2 branch 2 times, most recently from 1bb4e16 to 5732a7d Compare September 13, 2024 14:53

vitlibar removed the do not test disable testing on pull request label Sep 13, 2024

vitlibar force-pushed the restore-access-dependencies-2 branch 7 times, most recently from 5bae8f5 to 4d5859e Compare September 14, 2024 07:07

qoega marked this pull request as ready for review September 16, 2024 08:40

qoega added pr-must-backport Pull request should be backported intentionally. Use this label with great care! pr-must-backport-cloud labels Sep 16, 2024

vitlibar force-pushed the restore-access-dependencies-2 branch 2 times, most recently from eb5c7b1 to e44e2c9 Compare September 19, 2024 07:53

vitlibar requested a review from pufit September 19, 2024 07:53

vitlibar assigned pufit Sep 19, 2024

vitlibar added 4 commits September 19, 2024 12:31

Refactoring of the code for making a backup and restoring access enti…

3c4d650

…ties.

Add some logging to restoring access entities.

91a91be

Remove unresolved dependencies while restoring with "allow_unresolved…

6b461fc

…_access_dependencies = true".

Add UUID to UpdateFunc in IAccessStorage.

64b7e2d

vitlibar force-pushed the restore-access-dependencies-2 branch from e44e2c9 to 9e35eb6 Compare September 19, 2024 10:31

vitlibar added 2 commits September 19, 2024 12:43

Add new restore setting "update_access_entities_dependents" to allow …

9247b32

…updating dependents of restored access entities.

Add more tests.

94f4b4d

vitlibar added 3 commits September 19, 2024 12:43

Change filenames of files with access entities inside a backup:

712a726

use "access-<UUID>.txt" instead of "access01.txt", "access02.txt", ...

Extend IAccessStorage::findAll() to enable finding all the entities i…

1a13355

…n the storage.

Support writing dependents of access entities to backup, added backup…

f8aec64

… setting "write_access_entities_dependents".

vitlibar force-pushed the restore-access-dependencies-2 branch from 9e35eb6 to 443ff06 Compare September 19, 2024 10:43

vitlibar added 5 commits September 19, 2024 19:28

Make restore setting "skip_unresolved_access_dependencies" work for d…

2d186bf

…ependents too.

Rename restore setting "allow_unresolved_access_dependencies" -> "ski…

d34b509

…p_unresolved_access_dependencies".

Enable by default backup setting "write_access_entities_dependents".

f2a0dd9

Enable by default restore setting "update_access_entities_dependents".

6bd7a74

Add one more test.

8a7a411

vitlibar force-pushed the restore-access-dependencies-2 branch from 443ff06 to 8a7a411 Compare September 19, 2024 17:30

pufit approved these changes Sep 26, 2024

View reviewed changes

alexey-milovidov added this pull request to the merge queue Sep 28, 2024

Merged via the queue into ClickHouse:master with commit cbfbdbf Sep 28, 2024

robot-ch-test-poll2 added the pr-synced-to-cloud The PR is synced to the cloud repo label Sep 28, 2024

robot-clickhouse added a commit that referenced this pull request Sep 28, 2024

Backport #69563 to 24.9: Improve restoring of access entities' depend…

5579d9f

…encies #2

robot-ch-test-poll3 mentioned this pull request Sep 28, 2024

Backport #69563 to 24.9: Improve restoring of access entities' dependencies #2 #70100

Merged

robot-ch-test-poll4 added the pr-backports-created Backport PRs are successfully created, it won't be processed by CI script anymore label Sep 28, 2024

robot-ch-test-poll3 added a commit that referenced this pull request Sep 28, 2024

Merge pull request #70100 from ClickHouse/backport/24.9/69563

a08a8ee

Backport #69563 to 24.9: Improve restoring of access entities' dependencies #2

vitlibar deleted the restore-access-dependencies-2 branch November 27, 2024 21:02

robot-ch-test-poll1 added the pr-backports-created-cloud deprecated label, NOOP label Nov 27, 2024

Conversation

vitlibar commented Sep 12, 2024

Changelog category:

Changelog entry:

Uh oh!

robot-clickhouse commented Sep 12, 2024 • edited by robot-ch-test-poll2 Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

qoega commented Sep 16, 2024

Uh oh!

vitlibar commented Sep 16, 2024

Uh oh!

vitlibar commented Sep 23, 2024

Uh oh!

pufit left a comment

Choose a reason for hiding this comment

Uh oh!

pufit Sep 26, 2024

Choose a reason for hiding this comment

Uh oh!

vitlibar Sep 26, 2024

Choose a reason for hiding this comment

Uh oh!

pufit Sep 26, 2024

Choose a reason for hiding this comment

Uh oh!

vitlibar Sep 26, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pufit Sep 26, 2024

Choose a reason for hiding this comment

Uh oh!

vitlibar Sep 26, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pufit Sep 26, 2024

Choose a reason for hiding this comment

Uh oh!

pufit Sep 26, 2024

Choose a reason for hiding this comment

Uh oh!

vitlibar commented Sep 28, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

alexey-milovidov commented Sep 28, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

robot-clickhouse commented Sep 12, 2024 •

edited by robot-ch-test-poll2

Loading

vitlibar Sep 26, 2024 •

edited

Loading

vitlibar Sep 26, 2024 •

edited

Loading

vitlibar commented Sep 28, 2024 •

edited

Loading