danny0405 commented on code in PR #11559: URL: https://github.com/apache/hudi/pull/11559#discussion_r1664978373
########## rfc/rfc-80/rfc-80.md: ########## @@ -0,0 +1,161 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> +# RFC-80: Support column families for wide tbles + +## Proposers + +- @xiarixiaoyao +- @wombatu-kun + +## Approvers + - + - + +## Status + +JIRA: https://issues.apache.org/jira/browse/HUDI- + +## Abstract + +In streaming processing, there are often scenarios where the table is widened. The current mainstream real-time stretching is completed through Flink's multi-layer join; +Flink's join will cache a large amount of data in the state backend. As the data set increases, the pressure on the Flink task state backend will gradually increase, and may even become unavailable. +In multi-layer join scenarios, this problem is more obvious. + +## Background +Currently, Hudi organizes data according to fileGroup granularity. The fileGroup is further divided into column clusters to introduce the columnFamily concept. +The organizational form of Hudi files is divided according to the following rules: +The data in the partition is divided into buckets according to hash; the files in each bucket are divided according to columnFamily; multiple colFamily files in the bucket form a completed fileGroup; when there is only one columnFamily, it degenerates into the native Hudi bucket table. + +![table](table.png) + +After splitting the fileGroup by columnFamily, the naming rules for base files and log files change. We add the cfName suffix to all file names to facilitate Hudi itself to distinguish column families. The addition of this suffix is compatible with Hudi's original naming method and has no conflict. + +![filenames](filenames.png) + +## Implementation +Describe the new thing you want to do in appropriate detail, how it fits into the project architecture. +Provide a detailed description of how you intend to implement this feature.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. + +### Constraints and Restrictions +1. The overall design relies on the lock-free concurrent writing feature of Hudi 1.0. Review Comment: It's not lock-free, it's just non-blocking, Hudi table utilies the lock to keep the instant time generation monotonically increasing. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@hudi.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org