The majority of key applications within IT Services and across the university store information in some form of database. For example, many of the commonly deployed web-based applications such as MediaWiki, Drupal, and WordPress require a MySQL database. IT Services must continue to support scalable, secure, and reliable MySQL, Microsoft SQL, and Oracle database platforms for the array of enterprise and department applications at Stanford.
Technologies in this section
Goals
- Provide better performance, horizontal scalability, and automated fail-over capability, particularly for the MySQL service.
- Explore a fee-based MySQL service.
- Inclusion of Business Affairs database applications in outsourced management service.
- Explore the creation of an MySQL environment for Restricted Data.
- Add flexibility to create different type of database accounts with different privileges, or let the users control the database operations.
- Expand the MySQL service to all full SUNetID accounts holders.
- Provide a better web-based management interface to users of the MySQL service.
Roadmap
- Adoption of Stanford's standards and practices for authentication, access control, change management, and incident reporting by IT Services' external partner.
- Demonstrate value of managed service to university leadership.
- Expansion of the managed service to include a full-service MySQL offering for those clients who require additional support.
- Explore improvements in performance, high availability, scalability: local storage, possibly scaling out and using multiple slaves for read-only applications.
- Provide statistics and reports for capacity planning and trending: use of database storage space, use of memory engines.
- Consider server consolidation (e.g., merge mysql-user and mysql-fb into one server using multiple mysql instances).
Measures of success
- Reduction in frequency and duration of incidents.
- Continued reduction in client rates for Oracle and Microsoft SQL database: An increase in the number of databases under contract results in a subsequent reduction in the rate per database.
- Increased use of centrally maintained database services.
