IMPALA-10505: Avoid creating misleading audit logs

Before this patch, if a requesting user granted the privilege on a view
does not have the privilege on the table(s) on which the view is based,
only an audit log entry indicating a failed authorization with respect
to an underlying table will be produced, whereas the requesting user is
actually able to fetch the data from the view since the user is granted
the privilege to do so. Such an audit log entry, however, is misleading
and thus should not be produced at all. Moreover, the audit log entry
corresponding to the successful authorization with respect to the view
should also be created.

Recall that to authorize a query involving a view, Impala performs
privilege checks for both the view as well as the underlying table(s).
Thus, for a user granted the privilege on the view but not the
underlying tables, the privilege check for the view would succeed but
those for the underlying table(s) would fail. Each privilege check
results in an audit log entry produced by Ranger. These audit log
entries will be collected by Impala and will be sent back to Ranger
after the query authorization. In the case where there is at least one
AuthzAuditEvent indicating a failed privilege check, only the
AuthzAuditEvent corresponding to the first failed check will be sent
back to Ranger. Refer to RangerBufferAuditHandler#flush() for further
details. Impala performs checks for both the view as well as the
underlying table(s) so that it is able to disallow the requesting user
from accessing the runtime profile or execution summary when the
requesting user is not granted the privilege on the underlying table(s).
Note that allowing the requesting user the access to the runtime profile
would reveal the existence of the underlying tables.

This patch resolves the issue by specifying whether or not we should
retain the audit log entries when calling
BaseAuthorizationChecker#authorizePrivilegeRequest() so that Impala will
not collect the audit log entries resulting from the privilege checks
for the underlying table(s) of a view.

Testing:
 - Added new FE tests to verify that the correct audit log entry is
   produced after this patch.
 - Added a new E2E test to verify that a user not granted the privilege
   on the underlying table(s) of a view is still not able to access the
   runtime profile or execution summary even though the user is granted
   the privilege on the view.
 - Verified that the patch passes the core tests in the DEBUG build.

Change-Id: I02f40eb96d6ed863cd2cd88d717c354dc351a64c
Reviewed-on: http://gerrit.cloudera.org:8080/17078
Reviewed-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
Tested-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
5 files changed
tree: 249eca16a5ced55585a3021c1c98b32615dd5a01
  1. be/
  2. bin/
  3. cmake_modules/
  4. common/
  5. docker/
  6. docs/
  7. fe/
  8. infra/
  9. java/
  10. lib/
  11. security/
  12. shell/
  13. ssh_keys/
  14. testdata/
  15. tests/
  16. www/
  17. .clang-format
  18. .clang-tidy
  19. .gitattributes
  20. .gitignore
  21. buildall.sh
  22. CMakeLists.txt
  23. EXPORT_CONTROL.md
  24. LICENSE.txt
  25. LOGS.md
  26. NOTICE.txt
  27. README-build.md
  28. README.md
  29. setup.cfg
README.md

Welcome to Impala

Lightning-fast, distributed SQL queries for petabytes of data stored in Apache Hadoop clusters.

Impala is a modern, massively-distributed, massively-parallel, C++ query engine that lets you analyze, transform and combine data from a variety of data sources:

  • Best of breed performance and scalability.
  • Support for data stored in HDFS, Apache HBase, Apache Kudu, Amazon S3, Azure Data Lake Storage, Apache Hadoop Ozone and more!
  • Wide analytic SQL support, including window functions and subqueries.
  • On-the-fly code generation using LLVM to generate lightning-fast code tailored specifically to each individual query.
  • Support for the most commonly-used Hadoop file formats, including Apache Parquet and Apache ORC.
  • Support for industry-standard security protocols, including Kerberos, LDAP and TLS.
  • Apache-licensed, 100% open source.

More about Impala

The fastest way to try out Impala is a quickstart Docker container. You can try out running queries and processing data sets in Impala on a single machine without installing dependencies. It can automatically load test data sets into Apache Kudu and Apache Parquet formats and you can start playing around with Apache Impala SQL within minutes.

To learn more about Impala as a user or administrator, or to try Impala, please visit the Impala homepage. Detailed documentation for administrators and users is available at Apache Impala documentation.

If you are interested in contributing to Impala as a developer, or learning more about Impala's internals and architecture, visit the Impala wiki.

Supported Platforms

Impala only supports Linux at the moment. Impala supports x86_64 and has experimental support for arm64 (as of Impala 4.0). Impala Requirements contains more detailed information on the minimum CPU requirements.

Export Control Notice

This distribution uses cryptographic software and may be subject to export controls. Please refer to EXPORT_CONTROL.md for more information.

Build Instructions

See Impala's developer documentation to get started.

Detailed build notes has some detailed information on the project layout and build.