Query Interceptors should not impact Spark usage

Query Interceptors should not impact Spark usage

While using a Query Interceptor, I noticed that it impacted Spark usage where the interceptor was not available on the classpath.

Original JIRA Issue: https://geomesa.atlassian.net/browse/GEOMESA-2892

Key: GEOMESA-2892
Type: Bug
Priority: Major
Status: To Do
Resolution: Unresolved
Reporter: James Hughes
Created: Thu, 16 Jul 2020 11:07:08 -0400
Updated: Wed, 30 Sep 2020 13:18:45 -0400
Affects Version(s): 3.0.0

Comments:

557058:da7f6dfa-f0a1-468b-aec5-2cb61273feb5 (Mon, 27 Jul 2020 16:49:36 -0400):

How does it impact spark usage? it should use a try/catch when loading query interceptors, so if it's missing it should continue on without it.

557058:da7f6dfa-f0a1-468b-aec5-2cb61273feb5 (Wed, 2 Sep 2020 16:27:38 -0400):

James Hughes any more details here? it's wrapped in a try/catch so i don't think it should affect spark if it's missing.

557058:48912df4-0004-4f90-9915-db26684ec006 (Thu, 3 Sep 2020 12:36:09 -0400):

It may have just been very noisy… This ticket can be low priority until we hit the issue again…

No source file, test, or entry point is identified. Start by reproducing Spark usage with a Query Interceptor absent from the classpath, then trace interceptor loading and its error handling; done means Spark continues to work without the interceptor and the behavior is covered by a regression test.