Fix flaky test in ElasticJobExecutorTest

[bee0511]

Fixes #2485 .

Changes proposed in this pull request:

• Modified getErrorMessage to isValidErrorMessage() method in ElasticJobExecutorTest to accept both possible orderings of ConcurrentHashMap entries ({0=..., 1=...} or {1=..., 0=...})
• This fix eliminates the flaky test failure caused by non-deterministic HashMap iteration order

[linghengqian]

• The code style is handled in a separate Maven profile. The PR did not pass the code formatting validation. You can execute first ./mvnw spotless:apply -Pcheck -T1C. Then execute ./mvnw checkstyle:check -Pcheck -T1C to manually adjust to repair the CI.

[bee0511]

• The code style is handled in a separate Maven profile. The PR did not pass the code formatting validation. You can execute first ./mvnw spotless:apply -Pcheck -T1C. Then execute ./mvnw checkstyle:check -Pcheck -T1C to manually adjust to repair the CI.

Hi @linghengqian,
I have fixed the checkstyle issue and the CI has passed.

[linghengqian]

LGTM