Abstract—To realize authentication and access control of multi-tenant is one of the important problems to build big data platform. This thesis raises an access control policy based on the authentication mechanism of traditional Kerberos and the new incubation project of Apache, Apache Sentry. By combining Sentry and traditional protocols and the advantages of Sentry in the ecology of Hadoop, authentication and access control of big data platform are completed to guarantee the strong isolation among different tenants and moderate isolation among resources of different tenants.
Index Terms—Multi-tenant, big data platform, Kerberos, apache sentry, Hadoop.
Yuanyuan Chang, Meina Song, and Haihong E. are with the School of Computer Science, Beijing University of Posts and Telecommunications, Beijing, China (e-mail: cyybupt@126.com, {mnsong, ehaihong}@bupt.edu.cn).
[PDF]
Cite:Yuanyuan Chang, Meina Song, and Haihong E, "Multi-Tenant Management and Research Based on the Big Data Platform of Sentry," International Journal of Computer Theory and Engineering vol. 9, no. 6, pp. 433-437, 2017.
