Changes to API for 1.5, you're probably best person for reviewing

The code changes look sane.

As I don't know what got changed with ColumnStore 1.5 compared to ColumnStore 1.2 I can't really assess the functionality impacts of these changes. This would require in-depth testing for every supported OS. The test suites should be a good indicator if everything is working as expected.

In my brief glance through the changes I only found following issues:
1. The spark-connector docker-compose.yaml files haven't been upgraded to multi node ColumnStore 1.5 [cf. 1, 2]
2. The Kettle data adapters tests still contain `COLUMNSTORE_INSTALL_DIR` instead of `COLUMNSTORE_XML_DIR`

jens@drop-table:~/mariadb-columnstore-data-adapters$ grep -rl "COLUMNSTORE_INSTALL_DIR" .

./kettle-columnstore-bulk-exporter-plugin/README.md

./kettle-columnstore-bulk-exporter-plugin/test/tests/all-datatype-ingestion-test/job.parameter.win

./kettle-columnstore-bulk-exporter-plugin/test/tests/cs-block-test/job.parameter.win

./kettle-columnstore-bulk-exporter-plugin/test/tests/csv-bool-test/job.parameter.win

./kettle-columnstore-bulk-exporter-plugin/test/tests/csv-ingestion-test/job.parameter.win

./kettle-columnstore-bulk-exporter-plugin/test/tests/csv-timestamp-test/job.parameter.win

./README.md

I also noted that the list of supported operating systems didn't change compared to 1.2.
It might be worth to revise if we still want to support the deprecated Debian 8 or want to introduce support for Debian 10.
The same goes for CentOS 8 and Ubuntu 20.04.

I also noted that the API still supports the deprecated Python 2 version. It might be worth to revise this as well.

To check the functionality properly it would require in-depth testing. I would assume 1-2 working days should be sufficient to validate the functionality. I unfortunately don't have any engineering cycles left in this sprint to do such testing. Therefore, I can't make any statements of the functional state of the data-adapters or api. I talked to one of my managers and they would be willing to consider it in our next sprint planning if required.

[1] https://github.com/mariadb-corporation/mariadb-columnstore-api/blob/MCOL-4101/spark-connector/python/docs/install-environment/docker-compose.yml#L70
[2] https://github.com/mariadb-corporation/mariadb-columnstore-api/blob/MCOL-4101/spark-connector/scala/docs/install-environment/docker-compose.yml#L70

Hi ben.thompson, I just found this ticket in my list.
I think the Jira processes differ between the CS and SkySQL team.
Do you expect me to do anything else, like to merge the PR or to move the ticket to test or something?

The same would apply to MCOL-4041

Do we need this anymore? Seems this is passed up by events. Backup/Restore is a new project.