KAFKA-15853: Move configDef out of core#16116
Conversation
|
@OmniaGM Could you please fix the conflicts |
51c53bd to
3e580ce
Compare
3e580ce to
3c70661
Compare
753ecbb to
1a765a5
Compare
|
@OmniaGM Please fix the conflicts, thanks! |
|
@OmniaGM Sorry that please fix the conflicts again :_ |
72c3776 to
9aef77c
Compare
|
There are more conflicts now. @OmniaGM Can you, please, fix it. |
|
Fixed let's hope this is the last one :) |
if it can go conflicted, it will :) |
| import static org.apache.kafka.common.config.ConfigDef.Type.SHORT; | ||
| import static org.apache.kafka.common.config.ConfigDef.Type.STRING; | ||
|
|
||
| public class KafkaConfig { |
There was a problem hiding this comment.
Could you please add comments to explain why we have "two" KafkaConfig in code base? That can avoid confusion when someone try to add something to KafkaConfig in the future.
- Make org.apache.kafka.server.config.KafkaConfig abstract so others can start move their methods there
chia7712
left a comment
There was a problem hiding this comment.
@OmniaGM thanks for updated PR. could you please fix build error?
[2024-06-10T15:13:31.280Z] > Task :core:compileScala
[2024-06-10T15:13:31.280Z] [Error] /home/jenkins/jenkins-agent/workspace/Kafka_kafka-pr_PR-16116/core/src/main/scala/kafka/server/KafkaConfig.scala:27:40: Unused import
| * Any code depends on kafka.server.KafkaConfig will keep for using kafka.server.KafkaConfig for the time being until we move it out of core | ||
| * For more details check KAFKA-15853 | ||
| */ | ||
| abstract public class KafkaConfig extends AbstractConfig { |
There was a problem hiding this comment.
It seems public abstract class is used more frequent than abstract public class. Maybe we can align it?
|
@OmniaGM Sorry that conflicts happen again :_ |
Done let's hope this is the final one :D |
mimaison
left a comment
There was a problem hiding this comment.
LGTM
I left a couple of suggestions but these can be addressed in one of the follow up PRs you have for KafkaConfig
mimaison
left a comment
There was a problem hiding this comment.
LGTM
I left a couple of suggestions but these can be addressed in one of the follow up PRs you have for KafkaConfig
|
I have renamed the class to |
Looking into this again. I think we can squeeze part of this vision in as it will reduce conflict with this work going forward specially with the ongoing work for shared groups and new coordinator which I want to avoid as it is hardest to fix. I published this in 88122cb @chia7712 and @mimaison please have a look again. I would really appreciate if we can merge this pr as it is getting painful to keep up with the conflicts :) For future improvements we can see the possibility of moving getters/ and validations logic as well into these specific config classes but not in the scope of KAFKA-15853. |
|
@OmniaGM |
7c173f7 to
5389335
Compare
I love this code style
I've set the alarm for this PR :) |
|
wait a minute. it seems there are failed tests related to this PR https://ci-builds.apache.org/blue/organizations/jenkins/Kafka%2Fkafka-pr/detail/PR-16116/19/tests |
mimaison
left a comment
There was a problem hiding this comment.
This is a bit of a weird state but I understand this is temporary and you have follow up PRs that should tidy things up.
LGTM, let's get this merged as it's a pain to keep in sync
|
None of the test failures are related, merging to trunk |
Moving configDef from core into
server::KafkaConfig, keeping the old reference fromcore::KafkaConfigfor the time being to avoid the confusion when we have need to importserver::KafkaConfigandcore::KafkaConfig.The following PRs will move the rest of core::KafkaConfig`
This will reduce conflicts while finishing the last part of moving KafkaConfig out of core
Committer Checklist (excluded from commit message)