[X] I have searched in the issues and found no similar issues.
What would you like to be improved?
In the Optimizing Plan phase, it is necessary to obtain the file's PartitionSpec. However, calling getMixedTablePartitionSpecById in a loop is a very expensive and unnecessary operation。
Search before asking
What would you like to be improved?
In the Optimizing Plan phase, it is necessary to obtain the file's PartitionSpec. However, calling
getMixedTablePartitionSpecById
in a loop is a very expensive and unnecessary operation。https://github.com/apache/amoro/blob/678b43c85347eb69d61e4f7ca016cb63d2ae56e4/amoro-ams/src/main/java/org/apache/amoro/server/optimizing/plan/OptimizingEvaluator.java#L119-L124
Especially in environments where Kerberos is enabled, repeatedly calling
org.apache.iceberg.BaseTable#specs
can incur lock overhead.How should we improve?
I want to place the code for obtaining PartitionSpec inside the TableFileScanHelper interface。
Are you willing to submit PR?
Subtasks
No response
Code of Conduct