KAFKA-19479: at_least_once mode in Kafka Streams silently drops messages when the producer fails with MESSAGE_TOO_LARGE, violating delivery guarantees

[shashankhs11]

Added 3 more tests to confirm existing behavior:

• Test testSplitAndReenqueue() for a single large record
• Test the retry behaviour of testSplitAndReenqueue() while splitting a single large record
• test hasRoomFor()

I also added a test for hasRoomFor()

• It always allows the first record no matter what the size is -- This is buggy behaviour. I think we should be checking the size but not sure why we always allow the first record.
• But after this, the numRecords become greater than 0, so it does not allow the subsequent records -- which is correct behaviour I suppose

All the assert statements are written to pass the tests confirming the existing buggy behaviour.

[github-actions]

A label of 'needs-attention' was automatically added to this PR in order to raise the
attention of the committers. Once this issue has been triaged, the triage label
should be removed to prevent this automation from happening again.