shivadarshan-devadiga commented on issue #8488:
URL: https://github.com/apache/gluten/issues/8488#issuecomment-6845782020

   @ayushi-agarwal I think this is fixed on main. Since 798ddcc54 ("[CORE][UI] 
Fix fallback info for V2 writes and align plan with Spark SQL tab", #11853), 
`GlutenQueryExecutionListener` posts a complete `GlutenPlanFallbackEvent` at 
the end of each execution, built by 
`GlutenImplicits.collectQueryExecutionFallbackSummary`. That path already walks 
into `InMemoryTableScanExec.relation.cachedPlan`, and the SQL UI stores this 
event for the execution.
   
   I checked it on current main (Velox, Spark 3.5) by caching a query whose 
Sort falls back (`spark.gluten.sql.columnar.sort=false`) and then reading the 
cache:
   
   ```scala
   val cached = spark.sql("SELECT c1, c2 FROM tmp1 ORDER BY c2").cache()
   cached.selectExpr("c1 + 1").collect()
   ```
   
   The last fallback event for that execution, the one shown in the UI, has the 
cached plan's fallback:
   
   ```
   numFallbackNodes=1 reasons=Map(009 Sort -> [FallbackByUserOptions] 
Validation failed on node Sort)
   ```
   
   So the Fallback Summary now includes fallbacks inside the cached plan and I 
think this issue can be closed.
   


-- 
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]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to