Enable Online DDL foreign key support (also in vtgate stress tests) when backing MySQL includes appropriate patch

[shlomi-noach]

Description

As explained in https://vitess.io/blog/2021-06-15-online-ddl-why-no-fk/, Online DDL cannot work with vanilla MySQL.

However, this patch, as part of this public fork, modifies MySQL as follows:

• Adds a rename_table_preserve_foreign_key global/session boolean variable (defaults false/0 for backwards compatibility).
• When enabled, a RENAME TABLES statement that replaces two tables Online-DDL-style, pins any children tables' foreign keys to the table name rather than the table pointer. Meaning after the RENAME, the children point to the new table that is using the original table name, rather than migrating along with the old table. This supports Online DDL for parent tables.
• Regardless of the variable, foreign key checks are ignored for any table that uses an internal vitess name (i.e. Online DDL vreplication tables, GC tables), whether parent-side or child-side. It's as if FOREIGN_KEY_CHECKS=0 for those specific tables. This supports Online DDL for children tables.

In this PR:

• Online DDL executor looks for the rename_table_preserve_foreign_key variable. If present, that means Online DDL ALTER TABLE is possible in vitess strategy.
• Executor validates existence of the variable before embarking on a foreign-key enabled migration. It also sets rename_table_preserve_foreign_key to 1 for the cut-over process (MySQL needs to prove it supports the variable at the beginning and at the end of a migration).
• VTGate FK stress tests now enable Online DDL when the backing MySQL server supports rename_table_preserve_foreign_key.

Note that https://github.com/vitessio/vitess uses vanilla MySQL in its CI workflows. Therefore, the changes made to the endtoend tests in this PR won't actually run in GitHub CI for this repo. The changes are stress tested (as part of any pull request) inside PlanetScale (developers of this patch) using aforementioned MySQL patch.

As a reminder, one still must use --unsafe-allow-foreign-keys DDL strategy when affected tables participate in FK constraint relationship.

Related Issue(s)

• Experimental foreign keys tracking issue #11975

Checklist

• "Backport to:" labels have been added if this change should be back-ported
• Tests were added or are not required
• Did the new or modified tests pass consistently locally and on the CI
• Documentation was added or is not required

Deployment Notes

[vitess-bot]

Review Checklist

Hello reviewers! 👋 Please follow this checklist when reviewing this Pull Request.

General

• Ensure that the Pull Request has a descriptive title.
• Ensure there is a link to an issue (except for internal cleanup and flaky test fixes), new features should have an RFC that documents use cases and test cases.

Tests

• Bug fixes should have at least one unit or end-to-end test, enhancement and new features should have a sufficient number of tests.

Documentation

• Apply the release notes (needs details) label if users need to know about this change.
• New features should be documented.
• There should be some code comments as to why things are implemented the way they are.
• There should be a comment at the top of each new or modified test to explain what the test does.

New flags

• Is this flag really necessary?
• Flag names must be clear and intuitive, use dashes (-), and have a clear help text.

If a workflow is added or modified:

• Each item in Jobs should be named in order to mark it as required.
• If the workflow needs to be marked as required, the maintainer team must be notified.

Backward compatibility

• Protobuf changes should be wire-compatible.
• Changes to _vt tables and RPCs need to be backward compatible.
• RPC changes should be compatible with vitess-operator
• If a flag is removed, then it should also be removed from vitess-operator and arewefastyet, if used there.
• vtctl command output order should be stable and awk-able.

[dbussink]

This is great!

[shlomi-noach]

onlineddl_scheduler test failures are legit and due to how we now error out when a migration is requested to --unsafe-allow-foreign-keys without an appropriate backing MySQL capability.

fixed