...

This page is meant as a template for writing a KIP. To create a KIP choose Tools->Copy on this page and modify with your content and replace the heading with the next KIP number and a description of your issue. Replace anything in italics with your own description.

Status

Current state:   [Under Discussion"]

Discussion thread: here [Change the link from the KIP proposal email archive to your own email thread]

JIRA: here [Change the link from KAFKA-1 to your own ticket]

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

Motivation

Describe the problems you are trying to solve.

Public Interfaces

Briefly list any new interfaces that will be introduced as part of this proposal or any existing interfaces that will be removed or changed. The purpose of this section is to concisely call out the public contract that will come along with this feature.

A public interface is any change to the following:

• Binary log format

• The network protocol and api behavior

• Any class in the public packages under clientsConfiguration, especially client configuration

  • org/apache/kafka/common/serialization

  • org/apache/kafka/common

  • org/apache/kafka/common/errors

  • org/apache/kafka/clients/producer

  • org/apache/kafka/clients/consumer (eventually, once stable)

• Monitoring

• Command line tools and arguments

• Anything else that will likely break existing users in some way when they upgrade

When tiered storage is enabled on the cluster, Kafka brokers has to build the remote log metadata for all the partitions on node restart. The remote log metadata is built in asynchronous fashion and does not interfere with the broker startup path. Once the broker becomes leader for some partitions, it cannot handle the client requests to access remote storage until the metadata gets built for those partitions. Currently, we are returning a ReplicaNotAvailable exception back to the client so that it will retry after sometime.

ReplicaNotAvailableException is applicable when there is a reassignment is in-progress and kind of deprecated with the NotLeaderOrFollowerException (PR#8979). It's good to introduce an appropriate retriable exception for remote storage to denote that it is not ready to accept the client requests yet.

Public Interfaces

New Exception class:

package org.apache.kafka.common.errors;

/**
 * An exception that indicates remote storage is not ready to receive the requests yet.
 */
public class RemoteStorageNotReadyException extends RetriableException {
    public RemoteStorageNotReadyException(String message, Throwable cause) {
        super(message, cause);
    }

    public RemoteStorageNotReadyException(String message) {
        super(message);
    }

    public RemoteStorageNotReadyException(Throwable cause) {
        super(cause);
    }

    public RemoteStorageNotReadyException() {
    }
}

Proposed Changes

Describe the new thing you want to do in appropriate detail. This may be fairly extensive and have large subsections of its own. Or it may be a few sentences. Use judgement based on the scope of the change.

When the metadata is not ready, instead of returning ReplicaNotAvailableException in RemotePartitionMetadataStore, we will return the new RemoteStorageNotReadyException.

Compatibility, Deprecation, and Migration Plan

• Both ReplicaNotAvailable and RemoteStorageNotReady exceptions extend RetriableException class.
• If the client implementation explicitly checks for ReplicaNotAvailableException instead of RetriableException, then it will break the clients. Tiered storage was not production ready in 3.6, so we don't expect any specific client side changes.
• What impact (if any) will there be on existing users?
• If we are changing behavior how will we phase out the older behavior?
• If we need special migration tools, describe them here.
• When will we remove the existing behavior?

Test Plan

Describe in few sentences how the KIP will be tested. We are mostly interested in system tests (since unit-tests are specific to implementation details). How will we know that the implementation works as expected? How will we know nothing broke?

• Will cover the patch with unit tests.

Rejected Alternatives

If there are alternative ways of accomplishing the same thing, what were they? The purpose of this section is to motivate why the design is the way it is and not some other way.