Status

Current state: Draft

Discussion thread: here

JIRA: here

Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).

Motivation

In our docs, we make statements about the Java versions we support. In the Hardware and OS section, we also have a few statements about the operating systems supported, but we don't actually say
anything about the hardware. Even if Kafka is a JVM based software, some of our dependencies like rocksdb, zstd, etc rely on native code so they may not work on all environments.

For years, the Apache Kafka CI only performed tests on x86_64 hardware. Recently, we've received donation of additional platforms and started testing on aarch64 and ppc64le. A s390x machine has also been donated but we're not yet using it.

Public Interfaces

None

Proposed Changes

I propose:
1) Making a statement about the hardware platforms we officially support
With the infrastructure we have in place today, the statement currently is that we officially support x86_64.

2) Defining a process for onboarding new platforms
To support a new platform, the project first needs hardware. Either it is already available in the Apache CI, or it is donated. Once added to the CI, all unit and integration tests must be passing. If a platform satisfy these requirements, it will be marked as supported in the next minor release.

Compatibility, Deprecation, and Migration Plan

None

Rejected Alternatives

None