release-21.2: changefeedccl: Increase message size limits for kafka sink.

[blathers-crl]

Backport 1/1 commits from #76265 on behalf of @miretskiy.

/cc @cockroachdb/release

---

Sarama library, used by kafka sink, limits the maximum message
sizes locally. When those limits are exceeded, sarama library
returns confusing error message which seems to imply that the remote
kafka server rejected the message, even though this rejection happened
locally:
kafka server: Message was too large, server rejected it to avoid allocation error.

This PR addresses the problem by increasing sarama limits to 2GB
(max int32).

An alternative approach was to extend kafka_sink_config to specify
maximum message size. However, this alternative is less desirable.
For one, the user supplied configuration can run afoul other limits
imposed by sarama library (e.g. MaxRequestSize), so more configuration
option must be added. In addition, this really exposes very low level
implementation details in the sarama library -- something that we
probably should not do.

Fixes #76258

Release Notes (enterprise change): Kafka sink supports larger messages,
up to 2GB in size.

---

Release justification:

[blathers-crl]

Thanks for opening a backport.

Please check the backport criteria before merging:

• Patches should only be created for serious issues or test-only changes.
• Patches should not break backwards-compatibility.
• Patches should change as little code as possible.
• Patches should not change on-disk formats or node communication protocols.
• Patches should not add new functionality.
• Patches must not add, edit, or otherwise modify cluster versions; or add version gates.

If some of the basic criteria cannot be satisfied, ensure that the exceptional criteria are satisfied within.

• There is a high priority need for the functionality that cannot wait until the next release and is difficult to address in another way.
• The new functionality is additive-only and only runs for clusters which have specifically “opted in” to it (e.g. by a cluster setting).
• New code is protected by a conditional check that is trivial to verify and ensures that it only runs for opt-in clusters.
• The PM and TL on the team that owns the changed code have signed off that the change obeys the above rules.

Add a brief release justification to the body of your PR to justify this backport.

Some other things to consider:

• What did we do to ensure that a user that doesn’t know & care about this backport, has no idea that it happened?
• Will this work in a cluster of mixed patch versions? Did we test that?
• If a user upgrades a patch version, uses this feature, and then downgrades, what happens?

[cockroach-teamcity]

This change is