yihua commented on code in PR #12005: URL: https://github.com/apache/hudi/pull/12005#discussion_r1788718857
########## rfc/rfc-82/rfc-82.md: ########## @@ -0,0 +1,98 @@ +<!-- + 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-82: Concurrency control of schema evolution + +## Proposers + +- @jonvex +- @Davis-Zhang-Onehouse + +## Approvers +- @yihua +- @sivabalan + +## Status + +JIRA: https://issues.apache.org/jira/browse/HUDI-8221 + +## Abstract + +This RFC proposes to enhance Hudi's optimistic concurrency control mechanism to handle concurrent schema evolution scenarios. The current implementation does not adequately address conflicts that may arise when multiple transactions attempt to alter the table schema simultaneously. This enhancement aims to detect and prevent such conflicts, ensuring data consistency and integrity in multi-writer environments. + +## Background + +Hudi supports optimistic concurrency control to handle concurrent write operations. However, the existing implementation does not specifically account for schema evolution conflicts. In a multi-writer environment, it's possible for different clients to attempt schema changes concurrently, which can lead to inconsistencies if not properly managed. + +Schema evolution is a critical operation that can significantly impact data compatibility and query results. Uncontrolled concurrent schema changes can result in data inconsistencies, failed reads, or incorrect query results. By extending the concurrency control mechanism to handle schema evolution, we can prevent these issues and ensure a more robust and reliable data management system. + +| Scenario | Schema when Txn Start | Schema when Txn Validates | Schema Used by curr Txn | Should Conflict? | Schema used by Commit Metadata of curr Txn | Notes | +|----------|---------------------|--------------------------|-------------------|------------------|---------------------------------------|-------| +| 1 | Not exists | Not exists | S1 | No | S1 | Current txn is the first commit ever, conflict is impossible | +| 2 | Not exists | S1 | S1 | No | S1 | Second commit, no schema evolution | +| 3 | Not exists | S1 | S2 | Yes | N/A (throws exception) | No predefined schema, effectively concurrent schema definition | Review Comment: If S2 can be evolved from S1, is that still OK? -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
