Caideyipi opened a new pull request, #18116:
URL: https://github.com/apache/iotdb/pull/18116

   ## Summary
   - Skip inactive aligned measurements when flushing `AlignedWritableMemChunk` 
to TsFile.
   - Keep active schema columns mapped to their underlying TVList column 
indexes during both aligned flush paths.
   - Add a minimal Pipe history IT for receiver 
`enable_auto_create_schema=false` with unflushed aligned `delete timeseries` 
before after-batch history sync.
   
   ## Root cause
   `delete timeseries` on an aligned measurement removes it from the active 
measurement map and marks its TVList column deleted, but the flush writer still 
used the full historical `schemaList`. A history Pipe created before the 
memtable was flushed could therefore produce a TsFile whose metadata still 
contained the deleted measurement, without a corresponding mod file. The 
receiver then tried to verify/create a schema that should no longer exist, and 
`LoadTsFile` failed or retried forever when auto-create schema was disabled.
   
   ## Tests
   - `mvn spotless:check -pl iotdb-core/datanode,integration-test -P 
with-integration-tests`
   - `git diff --check`
   - `mvn -pl integration-test -am -P with-integration-tests -DskipUTs 
-Dit.test=IoTDBPipeReceiverAutoCreateDisabledIT#testAutoSplitHistoryTsFileWithUnflushedAlignedDeletionWhenReceiverAutoCreateSchemaDisabled
 -DfailIfNoTests=false -Dfailsafe.failIfNoSpecifiedTests=false verify` 
*(blocked before the IT by existing upstream local compile failure in 
`iotdb-thrift`: generated thrift sources import missing `javax.annotation` on 
JDK 17)*


-- 
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]

Reply via email to