pvary commented on pull request #1789:
URL: https://github.com/apache/iceberg/pull/1789#issuecomment-731365732


   > @pvary, looks great! Thanks for looking into the archive options. I think 
we may be able to do this more simply by archiving reports instead of creating 
new log files. The reports I've used have had stderr/stdout.
   
   Thanks for reviewing @rdblue!
   I have tried archiving reports as a first option. You can see the results 
here in the artifact: https://github.com/apache/iceberg/runs/1429922302
   
   I found them lacking because of 2 reasons:
   1. Hive often does not propagate the exception to the client (the theory is 
that infra errors should not be handled by the users), so Hive based tests 
produce the following exceptions without enough details to investigate (INFO 
logs contain the real Exception):
   ```
   org.apache.iceberg.mr.hive.TestHiveIcebergStorageHandlerWithCustomCatalog > 
testScanTable[fileFormat=PARQUET, engine=tez] FAILED
       java.lang.IllegalArgumentException: Failed to execute Hive query 'SELECT 
* FROM default.customers ORDER BY customer_id DESC': Error while processing 
statement: FAILED
   : Execution Error, return code 1 from 
org.apache.hadoop.hive.ql.exec.tez.TezTask
           Caused by:
           org.apache.hive.service.cli.HiveSQLException: Error while processing 
statement: FAILED: Execution Error, return code 1 from 
org.apache.hadoop.hive.ql.exec.tez.TezTask
   ```
   2. Having separate directories for listing the tests for every subproject 
does not help in identifying the failing test at glance
   
   For the 2nd option I have tried the 
[official](https://github.com/gradle/gradle/blob/v5.4.1/subprojects/docs/src/samples/testing/testReport/groovy/build.gradle)
 solution for aggregating the results, but it only does the aggregation if 
there is no test failure which is less than ideal 😄
   
   So that is why I have decided to backtrack to the "have a file for a 
subproject output and keep the original test output" solution.
   
   After my trip to the test output aggregation I feel that if we want to have 
aggregate test results then it should be done in a different PR and here we 
should concentrate on the archiving the logs


----------------------------------------------------------------
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:
[email protected]



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

Reply via email to