HDDS-2352. Client gets internal error instead of volume not found in secure cluster

[adoroszlai]

What changes were proposed in this pull request?

Let checkAccess propagate original OMException with VOLUME_NOT_FOUND result code instead of new one with INTERNAL_ERROR. This is similar to existing logic for buckets.

https://issues.apache.org/jira/browse/HDDS-2352

How was this patch tested?

Tested original steps to reproduce:

$ cd hadoop-ozone/dist/target/ozone-0.5.0-SNAPSHOT/compose/ozonesecure
$ docker-compose exec scm bash
$ kinit -kt /etc/security/keytabs/testuser.keytab testuser/scm@EXAMPLE.COM
$ ozone freon ockg -n 1 -t 1
...
2019-10-23 18:52:25,424 [main] INFO       - Creating Volume: vol1, with testuser/scm@EXAMPLE.COM as owner.
2019-10-23 18:52:25,542 [main] INFO       - Creating Bucket: vol1/bucket1, with Versioning false and Storage Type set to DISK and Encryption set to false
...
Successful executions: 1

Also ran ozonesecure acceptance test.

[dineshchitlangia]

+1 LGTM, pending CI.
Thanks @adoroszlai for reporting & fixing this.

[bharatviswa504]

+1 LGTM.

[bharatviswa504]

Thank You @adoroszlai for the fix and @dineshchitlangia for the review.

[adoroszlai]

Thanks @dineshchitlangia for the prompt review, and @bharatviswa504 for reviewing and merging it.