morningman opened a new pull request, #65133:
URL: https://github.com/apache/doris/pull/65133

   ## Proposed changes
   
   Fix the chronically-failing P1 regression test 
`nereids_syntax_p1/mv/aggregate/agg_sync_mv`, which fails in every recent 
branch-4.0 CI run.
   
   ### Root cause
   
   The second `streamLoad` block still pointed at the pre-move location after 
#52242 moved the data file `agg_mv_test.dat` into the `aggregate/` subdirectory 
(and added a correct initial load block):
   
   - `db "regression_test_nereids_syntax_p1_mv"` — missing the `_aggregate` 
suffix; no test creates this database.
   - `file "../agg_mv_test.dat"` — the `../` resolves to 
`data/nereids_syntax_p1/mv/agg_mv_test.dat`, which does not exist (the file 
lives under `.../mv/aggregate/`).
   
   The missing data file makes the stream load return an empty body, which 
`StreamLoadAction` swallows (`streamLoadToBe` catches the exception and returns 
`null`) and later surfaces as the misleading 
`java.lang.IllegalArgumentException: Text must not be null or empty` from 
`JsonSlurper.parseText`.
   
   This was masked by an earlier failure at `mv_sync50` (duplicate internal 
column name `__variance_samp_1`); once #64974 fixed that, the test progressed 
to this block and this stale-path bug became the active failure.
   
   ### Fix
   
   Align the second load block with the working initial load block (lines 
114-130):
   
   - `db` → `regression_test_nereids_syntax_p1_mv_aggregate`
   - `file` → `agg_mv_test.dat`
   
   Test-only change.
   
   🤖 Generated with [Claude Code](https://claude.com/claude-code)
   


-- 
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