[fix][client] Fix perf-producer get OOM with high publish latency

[codelipenghui]

Motivation

Fix the perf-producer to get direct memory OOM with high publish latency.
The root cause is the perf-producer without any memory limitation and pending
queue size limitation by default.

You can try to set the max pending queue size for a workaround.

Modifications

Use the default configuration of the PulsarClient instead of removing the memory
limitation in pulsar-perf.

We should also provide an option to specify the max memory for the perf tool.
But it will introduce new public API changes to the perf tool, so it should only be available
in the feature release. This PR is focused on the BUG fix to avoid confusing users
with OOM.

Does this pull request potentially affect one of the following parts:

If the box was checked, please highlight the changes

• Dependencies (add or upgrade a dependency)
• The public API
• The schema
• The default values of configurations
• The threading model
• The binary protocol
• The REST endpoints
• The admin CLI options
• The metrics
• Anything that affects deployment

Documentation

• doc
• doc-required
• doc-not-needed
• doc-complete