Logstash Web server returns 500 Internal Server Error on aarch64 linux nodes

[robbavey]

Logstash information:

Please include the following information:

1. Logstash version (e.g. bin/logstash --version)

Snapshot versions of Logtstash 7.16.0, 8.0.0 and 8.1.0

2. Logstash installation source (e.g. built from source, with a package manager: DEB/RPM, expanded from tar or zip archive, docker)
   Tested on docker and from command line
3. How is Logstash being run (e.g. as a service/service manager: systemd, upstart, etc. Via command line, docker/kubernetes)

JVM (e.g. java -version):

Bundled JVM

OS version (uname -a if on a Unix-like system):

Description of the problem including expected versus actual behavior:

When attempting to access any of the Logstash API, an internal server error is returned, with no error message in the logs.
However, running in debug, the following error line is output:

[2021-11-23T16:46:55,224][DEBUG][logstash.agent           ] 2021-11-23 16:46:55 -0500 Read: #<NoMethodError: undefined method `>=' for nil:NilClass>

This appears to be related to this jruby error - it only manifests on aarch64 nodes as far as I can tell.

Further testing has shown that downgrading to Puma to 4.x resolves the error, but it persists in all versions of 5.x. See https://logstash-ci.elastic.co/job/elastic+logstash+master+multijob-aarch64-docker-acceptance/flavor=full,nodes=aarch64/80/console for an example docker integration test using a downgraded version of puma

Steps to reproduce:

Requires aarch64 enabled nodes, such as t4g graviton node type on AWS

1. Run minimal pipeline
2. Try and access http://localhost:9600
3. Note 500: Internal Server error is returned.

Provide logs (if relevant):

[2021-11-23T16:46:55,224][DEBUG][logstash.agent           ] 2021-11-23 16:46:55 -0500 Read: #<NoMethodError: undefined method `>=' for nil:NilClass>

[robbavey]

cc @roaksoax

[robbavey]

Fixed by #13447