bug: DuckDB tests sometimes fail #14548
Description
There seems to be a race between flushing and reading rows in the latest version of the DuckDB plugin, leading to this occasional test failure:
--- FAIL: TestPlugin (5.50s)
testing_write_delete.go:142:
Error Trace: /home/runner/go/pkg/mod/github.com/cloudquery/plugin-sdk/v4@v4.12.5/plugin/testing_write_delete.go:142
/home/runner/go/pkg/mod/github.com/cloudquery/plugin-sdk/v4@v4.12.5/plugin/testing_write.go:147
Error: Not equal:
expected: int(10)
actual : int64(20)
Test: TestPlugin
Messages: unexpected amount of items after second delete stale
--- FAIL: TestPlugin/TestDeleteStale (1.[9](https://github.com/cloudquery/cloudquery/actions/runs/6532488275/job/17735758206#step:6:10)4s)
--- FAIL: TestPlugin/TestDeleteStale/All (1.92s)
testing.go:[14](https://github.com/cloudquery/cloudquery/actions/runs/6532488275/job/17735758206#step:6:15)90: test executed panic(nil) or runtime.Goexit: subtest may have called FailNow on a parent test
testing.go:1490: test executed panic(nil) or runtime.Goexit: subtest may have called FailNow on a parent test
(Note: expected is 10, actual is 20, we can ignore the int64/int difference)