[Bug] unstable case gp_explain

[avamingli]

Apache Cloudberry version

No response

What happened

https://github.com/apache/cloudberry/actions/runs/15015096147/job/42192042707

0s
Run # Search for regression.diffs recursively
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_explain.out /__w/cloudberry/cloudberry/src/test/regress/results/gp_explain.out
--- /__w/cloudberry/cloudberry/src/test/regress/expected/gp_explain.out	2025-05-[1](https://github.com/apache/cloudberry/actions/runs/15015096147/job/42192042707#step:18:1)4 01:06:12.981846295 -0700
+++ /__w/cloudberry/cloudberry/src/test/regress/results/gp_explain.out	2025-05-14 01:06:12.997846264 -0700
@@ -672,7 +671,7 @@
                      ->  WorkTable Scan on y (never executed)
                      ->  Materialize (never executed)
                            ->  Gather Motion 3:1  (slice1; segments: 3) (never executed)
-                                 ->  Seq Scan on recursive_table_ic (actual rows=4061 loops=1)
+                                 ->  Seq Scan on recursive_table_ic (actual rows=6090 loops=1)
 GP_IGNORE:(13 rows)
 
 DROP TABLE recursive_table_ic;

Found in PR #1098, succeed after a retry.

It seems a catalog statistic issue, I haven't seen it before.

What you think should happen instead

No response

How to reproduce

CI

Operating System

CI

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.

[avamingli]

Should be fixed by #1140