Removed ReceiveBufferSize and SendBufferSize to improve message rates

[MarcoZuiderwijkBS]

Proposed Changes

When connecting our application running in the Western Europe region to a RabbitMQ instance in the East-US region we got a very low message rate (12 messages per second). Message size is around 80kb. We also tested this with a Python (using pika 1.3.2) and a Java consumer (using amqp-client 5.16.0) which had a much higher message rate (> 500 messages per second). We also did a test with RabbitMQ version 5.2.0 and to our surprise this also had a much higher message rate.

My colleague Joey Bleeker investigated the differences between version 5.2.0 and 6.6.0 of the RabbitMQ Client and found this commit. We removed the setting for the ReceiveBufferSize and SendBufferSize. This gives the latest RabbitMQ.Client the same performance as the version 5.2.0. The previous changes seems to conflict with the TCP Window Scaling.

Types of Changes

What types of changes does your code introduce to this project?
Put an x in the boxes that apply

• Bug fix (non-breaking change which fixes issue #NNNN)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause an observable behavior change in existing systems)
• Documentation improvements (corrections, new content, etc)
• Cosmetic change (whitespace, formatting, etc)

Checklist

Put an x in the boxes that apply. You can also fill these out after creating
the PR. If you're unsure about any of them, don't hesitate to ask on the
mailing list. We're here to help! This is simply a reminder of what we are
going to look for before merging your code.

• I have read the CONTRIBUTING.md document
• I have signed the CA (see https://cla.pivotal.io/sign/rabbitmq)
• All tests pass locally with my changes
• I have added tests that prove my fix is effective or that my feature works
• I have added necessary documentation (if appropriate)
• Any dependent changes have been merged and published in related repositories

Further Comments

If this is a relatively large or complex change, kick off the discussion by
explaining why you chose the solution you did and what alternatives you
considered, etc.

[michaelklishin]

@MarcoZuiderwijkBS thank you for digging in.

As ff627be mentions, some workload may benefit from these settings, while your workload with messages much larger than average suffers.

Sounds like it should be a configurable settings and not something we flip back and forth.

At least several other clients allow you to "post-configure" a socket before it connects using
a lambda. I'd suggest something like that, with the default being "no configuration besides Nagle's algorithm".

[lukebakken]

Converting to draft just to ensure this isn't merged prematurely.

[lukebakken]

We should adopt what the Java client does to configure the underlying Socket:

https://github.com/rabbitmq/rabbitmq-java-client/blob/main/src/main/java/com/rabbitmq/client/SocketConfigurator.java

That way, if someone wants to adjust buffer sizes, they can. The only thing we should do is disable Nagle's algo.

cc @stebet on this just so he's in the loop.

[lukebakken]

@MarcoZuiderwijkBS I've merged your code into #1415. Since your PR is on the main branch I can't easily make modifications to it.

[pivotal-cla]

@MarcoZuiderwijkBS Please sign the Contributor License Agreement!

Click here to manually synchronize the status of this Pull Request.

See the FAQ for frequently asked questions.

[lukebakken]

@MarcoZuiderwijkBS https://www.nuget.org/packages/RabbitMQ.Client/6.7.0-rc.1

[pivotal-cla]

@MarcoZuiderwijkBS Thank you for signing the Contributor License Agreement!

[pivotal-cla]

@MarcoZuiderwijkBS Thank you for signing the Contributor License Agreement!