Adds support for CHANGE REPLICATION SOURCE TO statement

[dennisurtubia]

SUMMARY

Adds new mode for community.mysql.mysql_replication module.

This new mode adds support for MySQL CHANGE SOURCE REPLICATION TO statement.

ISSUE TYPE

• Feature Pull Request

COMPONENT NAME

mysql_replication

ADDITIONAL INFORMATION

MySQL CHANGE MASTER TO statement was deprecated. Doc

I didn´t found a respective statement for MariaDB that replaces CHANGE MASTER TO. List of MariaDB commands

Resolves #635

[dennisurtubia]

@Andersson007 CHANGE REPLICATION SOURCE TO doesn't uses PRIMARY nomenclature in the statement variables, instead it, uses SOURCE doc.

I kept the nomenclature so as not to significantly increase the list of parameters.

[dennisurtubia]

I don't like this similar repetition with changeprimary mode. I can refactor if this make sense.

[Andersson007]

@dennisurtubia hello, thanks for the PR! Could you please add a changelog fragment and ping us when it's ready for review after the CI is green, thanks!

[Andersson007]

@dennisurtubia yep, any meaningful refactoring is welcome but it'd be better to do it in a dedicated PR. And also feel free to add yourself to the authors: field in DOCUMENTATION if you're not already there.

[codecov]

Codecov Report

Attention: Patch coverage is 49.09091% with 28 lines in your changes are missing coverage. Please review.

Project coverage is 75.34%. Comparing base (47710cf) to head (6405ddc).
Report is 1 commits behind head on main.

❗ Current head 6405ddc differs from pull request most recent head 9028d2c

Please upload reports for the commit 9028d2c to get more accurate results.

Files	Patch %	Lines
plugins/modules/mysql_replication.py	49.09%	13 Missing and 15 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #636      +/-   ##
==========================================
- Coverage   76.38%   75.34%   -1.04%     
==========================================
  Files          19       18       -1     
  Lines        2456     2507      +51     
  Branches      618      640      +22     
==========================================
+ Hits         1876     1889      +13     
- Misses        403      420      +17     
- Partials      177      198      +21     

Flag	Coverage Δ
integration	74.63% <49.09%> (-0.54%)	⬇️
sanity	?

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Andersson007]

@dennisurtubia LGTM, one small formatting thing ^

[Andersson007]

@dennisurtubia thanks, LGTM

@laurent-indermuehle any thoughts?

[laurent-indermuehle]

I'm grumpy. First we had to rename master to primary and slave to replica. Almost done but there is still a lot of 'master' and 'slave' in many commands.
Then Oracle decide that primary is not good enough and switch to 'source'. Nice job. I'm not exhausted /s. Thanks.

Anyway, my only request about this PR is that the option is more clearly labeled "only for MySQL" please.

[dennisurtubia]

I'm grumpy. First we had to rename master to primary and slave to replica. Almost done but there is still a lot of 'master' and 'slave' in many commands. Then Oracle decide that primary is not good enough and switch to 'source'. Nice job. I'm not exhausted /s. Thanks.

Anyway, my only request about this PR is that the option is more clearly labeled "only for MySQL" please.

Hello, @laurent-indermuehle. I just updated the changereplication mode description, including "only supported in MySQL 8.0.23 and later"

[Andersson007]

@laurent-indermuehle can we merge the PR now?

[laurent-indermuehle]

Yes, Thanks @dennisurtubia having made the change I requested. I merge now! Thanks for your contribution!

[betanummeric]

Hi @dennisurtubia, thanks for the PR.
But variables should be put into SQL by the mysql driver, not by string replacement.
cursor.execute("select %(value)s", dict(value=1)) is safe,
whereas cursor.execute("select %(value)s" % dict(value=1)) allows SQL injection attacks.
Please fix this in a new PR to make the module safer.

[laurent-indermuehle]

@betanummeric Ho good catch, thanks for the advice.
What about this, is it unsafe too?

query = 'START %s %s' % (term, ','.join(chm))
cursor.execute(query)

[betanummeric]

@laurent-indermuehle Yes, this is unsafe as well. The cursor.execute function adds safety, the other is just concatenating strings without checking anything.

[Andersson007]

@dennisurtubia could you fix this? if no time, can someone else do it?

[laurent-indermuehle]

@betanummeric Then this whole file is unsafe: https://github.com/eryx12o45/community.mysql/blob/main/plugins/modules/mysql_replication.py

Because we keep track of the executed queries in a list, how to make the cursor.execute safe?
Do you think we need to open an issue for this?

[dennisurtubia]

@dennisurtubia could you fix this? if no time, can someone else do it?

Hey @Andersson007. Yes, I will fix it this week

[dennisurtubia]

@betanummeric Then this whole file is unsafe: https://github.com/eryx12o45/community.mysql/blob/main/plugins/modules/mysql_replication.py

Because we keep track of the executed queries in a list, how to make the cursor.execute safe? Do you think we need to open an issue for this?

@laurent-indermuehle I noticed it too. I used some existing code in mysql_replication module to write this function.

I think it would be necessary to create an issue to resolve all issues.

[Andersson007]

@dennisurtubia I've created the issue, please take a look