Improve docs on the ingress module

[mtopolnik]

Closes #61. Closes #60.

[amunra]

Clarification:

I had written feedback about reverting buffer -> buf changes. I don't particularly care which variable name is used, so long as it's consistent with the rest of the examples in questdb-rs/examples as well.

[mtopolnik]

I had a specific reason to use buf instead of buffer -- making the construct sender.flush(&mut buffer) a bit shorter, as I was going to use it in sentences whenever referring to flush. But in the end, I use it just once, so I'll revert to buffer.

[mtopolnik]

cargo test --no-default-features doesn't compile on either main or this branch.

[amunra]

Sorry. Correct. Need to enable at least webpki-roots.

[mtopolnik]

test_tls_ca_webpki_and_os_roots is failing with

LineProtocolException: [-1] table: 77836915f4494765acbd56ec18f6dd7e, column: a; cast error from protocol type: TAG to column type: STRING

My guess is that there's something wrong with the isolation between tests. I suppose this test should create its own table, but it seems to already exist and there's a type clash on column a.

[mtopolnik]

Working on the docs, I noticed that line_sender.hpp doesn't declare flush_and_keep_with_flags, so it seems you can't ensure a transactional flush from C++.

[jerrinot]

I think the Ingress doc is missing a chapter on error handling. there is only this:

Finally, the client will keep retrying the request if it experiences errors. You can configure the total time budget for retrying: retry_timeout (milliseconds, default 10 seconds)

A good understanding of error handling is critical for successful usage. Something like this: https://py-questdb-client.readthedocs.io/en/latest/sender.html#error-reporting Or here is Java: https://questdb.io/docs/clients/java_ilp/#error-handling