n3nash commented on issue #2802: URL: https://github.com/apache/hudi/issues/2802#issuecomment-818457749
@aditiwari01 What failure do you see ? Generally the following is the execution path 1. Provide new writer schema to Hudi 2. Write data with new writer schema to hudi log files Here there are 2 possibilities a. Commit this batch succeeds - In this case, the writerSchema for the `AbstractRealtimeRecordReader` will be this latest schema with which the latest data was written b. Commit this batch fails - In this case, the writerSchema for the `AbstractRealtimeRecordReader` will be this latest schema by the previous successful commit with which the latest data for that commit was written. This failed commit will not have the latest schema or the data. There shouldn't be a scenario where inside Hudi, a data record for the latest schema is written while the schema is not committed or vice-versa. If you see some exceptions, please paste it here so I can understand your problem in more details. -- 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
