Describe per-connector support for fault-tolerant execution

[jhlodin]

Description

Add newly-uncovered information about per-connector support for fault-tolerant execution.

Is this change a fix, improvement, new feature, refactoring, or other?

Improvement

Is this a change to the core query engine, a connector, client library, or the SPI interfaces? (be specific)

Documentation

How would you describe this change to a non-technical end user or system administrator?

Adds information necessary to understand how fault-tolerant execution behaves with different connectors

Related issues, pull requests, and links

Documentation

( ) No documentation is needed.
(x) Sufficient documentation is included in this PR.
( ) Documentation PR is available with #prnumber.
( ) Documentation issue #issuenumber is filed, and can be handled later.

Release notes

(x) No release notes entries required.
( ) Release notes entries required with the following suggested text:

# Section
* Fix some things. ({issue}`issuenumber`)

[mosabua]

This should be a warning imho and remove the "Note that" and maybe also do not quote the error message specifically

[jhlodin]

Switched to warning and reworded a bit to emphasize the "warning" part of this note.

I think the error message text is important to put here, because when troubleshooting people are going to search for that text. The text also does not mention fault-tolerant execution making it harder to trace the cause back here.

We'll just have to update the text when it changes, business as usual.

[mosabua]

I also think we need to figure out how document that these three connectors support FTE in the actual connector doc

[hashhar]

Please avoid using abbreviations in the PR title and commit messages - makes it harder to skim commits for people who don't know what they mean.

LGTM % Manfred's comments.