[Bug] CI: ic-cbdb-parallel failed

[tuhaihe]

Apache Cloudberry version

main

What happened

When the text changed PR cd451b9 was merged to the main, the ic-cbdb-parallel test failed：https://github.com/apache/cloudberry/actions/runs/13478582968/job/37661104121.

=== Starting results parsing for ic-cbdb-parallel ===
Number of configurations to parse: 1

=== Parsing results for configuration 1/1 ===
Make command: make -C src/test/regress installcheck-cbdb-parallel
Log file: build-logs/details/make-ic-cbdb-parallel-config0.log
Parsing test results from: build-logs/details/make-ic-cbdb-parallel-config0.log
Test Results:
Status: failed
Total Tests: 689
Failed Tests: 1
Ignored Tests: 0
Passed Tests: 688
Failed Test Names:
  - gp_aggregates
Test failures detected but properly parsed
MAKE_COMMAND="make -C src/test/regress installcheck-cbdb-parallel"
STATUS=failed
TOTAL_TESTS=689
PASSED_TESTS=688
FAILED_TESTS=1
IGNORED_TESTS=0
FAILED_TEST_NAMES=gp_aggregates
SUITE_NAME=ic-cbdb-parallel
DIR=src/test/regress
TARGET=installcheck-cbdb-parallel
Results stored in test_results.0.txt
=== End parsing for configuration 1 ===

=== Results file status ===
Generated results files:
- test_results.ic-cbdb-parallel.0.txt exists

Found regression.diffs at: ./src/test/regress/regression.diffs
diff -I HINT: -I CONTEXT: -I GP_IGNORE: -U3 /__w/cloudberry/cloudberry/src/test/regress/expected/gp_aggregates.out /__w/cloudberry/cloudberry/src/test/regress/results/gp_aggregates.out
--- /__w/cloudberry/cloudberry/src/test/regress/expected/gp_aggregates.out	2025-02-22 18:13:19.353471033 -0800
+++ /__w/cloudberry/cloudberry/src/test/regress/results/gp_aggregates.out	2025-02-22 18:13:19.365471163 -0800
@@ -1,11 +1,10 @@
 GP_IGNORE: formatted by atmsort.pm
 -- array_agg tests
 SELECT array_agg(a order by a) as a_by_a from aggtest;
-    a_by_a     
----------------
- {0,42,56,100}
-(1 row)
-
+DETAIL:  connection to server at "[17](https://github.com/apache/cloudberry/actions/runs/13478582968/job/37661104121#step:18:18)2.18.0.2", port 7002 failed: FATAL:  reader could not find writer proc entry
+DETAIL:  lock [0,1260] AccessShareLock 0. Probably because writer gang is gone somehow. Maybe try rerunning.
+ERROR:  failed to acquire resources on one or more segments
+ (seg0 172.[18](https://github.com/apache/cloudberry/actions/runs/13478582968/job/37661104121#step:18:19).0.2:7002)
 SELECT array_agg(b order by b) as b_by_b from aggtest;
            b_by_b            
 -----------------------------

What you think should happen instead

It should pass all the tests.

How to reproduce

See the test report.

Operating System

Rocky 9

Anything else

No response

Are you willing to submit PR?

• Yes, I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct.

[yjhjstz]

same with #837

[tuhaihe]

same with #837

Good to know. Another job failed again, and had the same error https://github.com/apache/cloudberry/actions/runs/13493540065/job/37696480485:

diff -I HINT: -I CONTEXT: -I GP_IGNORE: -U3 /__w/cloudberry/cloudberry/src/test/regress/expected/gp_sync_lc_gucs.out /__w/cloudberry/cloudberry/src/test/regress/results/gp_sync_lc_gucs.out
--- /__w/cloudberry/cloudberry/src/test/regress/expected/gp_sync_lc_gucs.out	2025-02-24 00:57:01.486[18](https://github.com/apache/cloudberry/actions/runs/13493540065/job/37696480485#step:18:19)4031 -0800
+++ /__w/cloudberry/cloudberry/src/test/regress/results/gp_sync_lc_gucs.out	2025-02-24 00:57:01.490184043 -0800
@@ -37,12 +37,10 @@
 ERROR:  terminating connection due to administrator command
 -- Should output the results given lc_monetary = 'en_US.utf8'
 SELECT to_char(c1, 'L9999999999999999.000')  FROM test_lc;
-        to_char         
-------------------------
- $ 45678901[23](https://github.com/apache/cloudberry/actions/runs/13493540065/job/37696480485#step:18:24)456789.000
- $-4567890123456789.000
-(2 rows)
-
+DETAIL:  connection to server at "172.18.0.2", port 7004 failed: FATAL:  reader could not find writer proc entry
+DETAIL:  lock [0,1[26](https://github.com/apache/cloudberry/actions/runs/13493540065/job/37696480485#step:18:27)0] AccessShareLock 0. Probably because writer gang is gone somehow. Maybe try rerunning.
+ERROR:  failed to acquire resources on one or more segments
+ (seg2 172.18.0.2:7004)
 RESET lc_monetary;
 -- Test if lc_time is synced
 SHOW lc_time;

[tuhaihe]

Hey @edespino could you please help take a look? Thanks.