Skip to content

Make the topic name configurable in storage paths#127

Open
Hariprasad Kuppuswamy (hariprasad-k) wants to merge 2 commits intoconfluentinc:masterfrom
hariprasad-k:master
Open

Make the topic name configurable in storage paths#127
Hariprasad Kuppuswamy (hariprasad-k) wants to merge 2 commits intoconfluentinc:masterfrom
hariprasad-k:master

Conversation

@hariprasad-k
Copy link

@hariprasad-k Hariprasad Kuppuswamy (hariprasad-k) commented Mar 5, 2020

Improvements to provide ability to exclude topic name in partition by configuration

@ghost
Copy link

Confluent Inc. (@confluentinc) It looks like Hariprasad Kuppuswamy (@hariprasad-k) just signed our Contributor License Agreement. 👍

Always at your service,

clabot

@hariprasad-k
Copy link
Author

Hariprasad Kuppuswamy (hariprasad-k) commented Mar 5, 2020

Fixed #126

@OneCricketeer

Question: What happens when you have topics as a list or topics.regex capture more than one topic?

Then all files in the directory will be of multiple topics?

public static final String TIMESTAMP_FIELD_NAME_DEFAULT = "timestamp";
public static final String TIMESTAMP_FIELD_NAME_DISPLAY = "Record Field for Timestamp Extractor";

public static final String TIMESTAMP_SCALING_FACTOR_CONFIG = "timestamp.scaling.factor";

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I suggest you break this out as a separate PR

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Jordan Moore (@OneCricketeer) Sure, I will break into separate PR.

@OneCricketeer

The patch also fixes #121 too

So does #122

@hariprasad-k
Copy link
Author

Hariprasad Kuppuswamy (hariprasad-k) commented Sep 9, 2020

Question: What happens when you have topics as a list or topics.regex capture more than one topic?

Then all files in the directory will be of multiple topics?

Jordan Moore (@OneCricketeer) Yes, all files of multiple topics will be in the same directory in this case.

But, this behavior will only happen if you use topics as a list or topics.regex in combination with path.include.topicname set as false. Further, as path.include.topicname is by default true, it will preserve the existing behavior by default.

@hariprasad-k
Copy link
Author

Hariprasad Kuppuswamy (hariprasad-k) commented Sep 9, 2020

Jordan Moore (@OneCricketeer) This PR does not have any changes related to Timestamp scaling factor anymore, which is covered by #160

Can you please review this changes ?

@OneCricketeer

I'm not a Confluent employee, so my reviews don't do anything

@dosvath Daniel (dosvath) changed the title Improvements to storage partitioning scheme to be flexible and configurable at runtime. Make the topic name configurable in storage paths Mar 2, 2021
@dosvath Daniel (dosvath) self-requested a review March 2, 2021 22:32
@dosvath
Copy link

Daniel (dosvath) commented Mar 2, 2021

Thanks Hariprasad Kuppuswamy (@hariprasad-k) for the contribution! This change will add flexibility to our connectors that use storage common to optionally exclude the topic name from the path (seems to be a popular request confluentinc/kafka-connect-hdfs#544). I'd like to get second view from Lev Zemlyanov (@levzem) before proceeding.

@hariprasad-k

Daniel (@dosvath) Lev Zemlyanov (@levzem) Any thoughts, or plans to review this contribution ?

@awebneck

Would be very much interested in having this feature integrated. The topic name being forcefully appended to the path limits us severely when sinking to already-established directory structures.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants