Edit D:\app\Administrator\product\11.2.0\dbhome_1\oc4j\j2ee\oc4j_applications\applications\em\em\online_help\tdpdw\tdpdw_security.htm
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"> <head> <meta http-equiv="Content-Type" content="text/html; charset=us-ascii" /> <meta http-equiv="Content-Style-Type" content="text/css" /> <meta http-equiv="Content-Script-Type" content="text/javascript" /> <title>Securing a Data Warehouse</title> <meta name="generator" content="Oracle DARB XHTML Converter (Mode = ohj/ohw) - Version 5.1" /> <meta name="date" content="2009-04-27T11:26:24Z" /> <meta name="robots" content="noarchive" /> <meta name="doctitle" content=" Securing a Data Warehouse" /> <meta name="relnum" content="11g Release 2 (11.2)" /> <meta name="partnum" content="E10578-01" /> <meta name="topic-id" content="TDPDW012" /> <link rel="copyright" href="./dcommon/html/cpyr.htm" title="Copyright" type="text/html" /> <link rel="stylesheet" href="./dcommon/css/blafdoc.css" title="Oracle BLAFDoc" type="text/css" /> <link rel="contents" href="toc.htm" title="Contents" type="text/html" /> <link rel="prev" href="tdpdw_bandr.htm" title="Previous" type="text/html" /> <script src="./callback.js" type="text/javascript"></script> <noscript>Your browser does not support JavaScript. This help page requires JavaScript to render correctly.</noscript> </head> <body> <div class="zz-skip-header"><a href="#BEGIN">Skip Headers</a></div> <table class="simple oac_no_warn" summary="" cellspacing="0" cellpadding="0" width="100%"> <col width="86%" /> <col width="*" /> <tr valign="bottom"> <td align="left"></td> <td align="center"><a href="tdpdw_bandr.htm"><img src="./dcommon/gifs/leftnav.gif" alt="Previous" /><br /> <span class="icon">Previous</span></a> </td> </tr> </table> <p><a id="CDDIFGGA" name="CDDIFGGA"></a><a id="TDPDW012" name="TDPDW012"></a></p> <h1><a name="TDPDW012|Basics of Data Warehouse Security"></a>Securing a Data Warehouse</h1> <p><a name="BEGIN" id="BEGIN"></a></p> <p>This section discusses consid<a id="sthref274" name="sthref274"></a><a id="sthref275" name="sthref275"></a>erations for data warehouse security, and includes the following topics:</p> <ul> <li> <p><a href="#CDDIFAIB">Overview of Data Warehouse Security</a></p> </li> <li> <p><a href="#CDDDIFGH">Using Roles and Privileges for Data Warehouse Security</a></p> </li> <li> <p><a href="#CDDIFGBA">Using Virtual Private Database in Data Warehouses</a></p> </li> <li> <p><a href="#CDDBADHH">Overview of Oracle Label Security</a></p> </li> <li> <p><a href="#CDDGEAGB">Overview of Fine-Grained Auditing in Data Warehouses</a></p> </li> <li> <p><a href="#CDDCEBAG">Overview of Transparent Data Encryption in Data Warehouses</a></p> </li> </ul> <p><a id="CDDIFAIB" name="CDDIFAIB"></a><a id="TDPDW0121" name="TDPDW0121"></a></p> <div class="sect1"><!-- infolevel="all" infotype="General" --> <h2><a name="TDPDW0121|Overview of Data Warehouse Security"></a>Overview of Data Warehouse Security</h2> <p>Data warehousing poses its own set of challenges for security. One major challenge is that enterprise data warehouses are often very large systems, serving many user communities with varying security needs. Thus, while data warehouses require a flexible and powerful security infrastructure, the security capabilities must operate in an environment that has stringent performance and scalability requirements.</p> <a id="TDPDW01211" name="TDPDW01211"></a> <div class="sect2"><!-- infolevel="all" infotype="General" --><a id="sthref276" name="sthref276"></a> <h3><a name="TDPDW01211|Why is Security Necessary for a Data Warehouse"></a>Why is Security Necessary for a Data Warehouse?</h3> <p>Many of the basic requirements for security are well-known, and apply equally to a data warehouse as they would to any other system: The applications must prevent unauthorized users from accessing or modifying data; the applications and underlying data must not be susceptible to data theft by hackers; the data must be available to the right users at the right time; and the system must keep a record of activities performed by its users.</p> <p>These requirements are perhaps even more important in a data warehouse because a warehouse contains data consolidated from multiple sources, and therefore, from the perspective of an individual trying to steal information, a data warehouse can be one of the most lucrative targets in an enterprise. In addition, a robust security infrastructure can often vastly improve the effectiveness or reduce the costs of a data warehouse environment.</p> <p>Some typical customer scenarios for data warehouse security include the following:</p> <ul> <li> <p>An enterprise is managing a data warehouse that will be widely used by many divisions and subsidiaries. This enterprise needs a security infrastructure that ensures the employees of each division be able to view only the data that is relevant to their own division, while also allowing employees in its corporate offices to view data for all divisions and subsidiaries.</p> </li> <li> <p>An enterprise's data warehouse stores personal information. Privacy laws may govern the use of such personal information. The data warehouse must handle data so as to adhere to these laws.</p> </li> <li> <p>An enterprise sells data from a data warehouse to its clients. Those clients may view only the data to which they have purchased or to which they have subscribed. They should never be permitted to see the data of other clients.</p> </li> </ul> </div> <!-- class="sect2" --></div> <!-- class="sect1" --> <p><a id="CDDDIFGH" name="CDDDIFGH"></a><a id="TDPDW0122" name="TDPDW0122"></a></p> <div class="sect1"><!-- infolevel="all" infotype="General" --> <h2><a name="TDPDW0122|Using Roles and Privileges for Data Warehouse Security"></a>Using Roles and Privileges for Data Warehouse Security</h2> <p>System <a id="sthref277" name="sthref277"></a><a id="sthref278" name="sthref278"></a>privileges, object privileges, and roles provide a basic level of database security. They are designed to control user access to data and to limit the kinds of SQL statements that users can execute. Roles are groupings of privileges that you can use to create different levels of database access. For example, you can create a role for application developers that enables users to create tables and programs.</p> <p>You can grant privileges and roles to other users only when you possess the necessary privilege. The granting of roles and privileges starts at the administrator level. At database creation, the administrative user <code>SYS</code> is created and granted all system privileges and predefined Oracle roles. User <code>SYS</code> can then grant privileges and roles to other users and also grant those users the right to grant specific privileges to others. Without explicitly granted privileges, a user cannot access any information in the database.</p> <p>Roles and privileges enforce security on the data itself, and their use is essential to a data warehouse because users access data through a number of applications and tools.</p> </div> <!-- class="sect1" --> <p><a id="CDDIFGBA" name="CDDIFGBA"></a><a id="TDPDW0123" name="TDPDW0123"></a></p> <div class="sect1"><!-- infolevel="all" infotype="General" --> <h2><a name="TDPDW0123|Using VPD in Data Warehouses"></a>Using Virtual Private Database in Data Warehouses</h2> <p>Virtual <a id="sthref279" name="sthref279"></a><a id="sthref280" name="sthref280"></a>private database (VPD) enables you to enforce security, to a fine level of granularity, directly on tables, views, or synonyms. Because security policies are attached directly to tables, views, or synonyms and automatically applied whenever a user accesses data, there is no way to bypass security. By dynamically appending SQL statements with a predicate, VPD limits access to data at the row level and ties a security policy to the database object itself. It enables multiple users to have secure direct access to critical data within a single database server, with the assurance of complete data separation. VPD can ensure that banking customers see only their own account history, and that an enterprise serving multiple companies' data (who may be competitors) can do so from the same data warehouse, and allow each company to see only its own data. In addition to control at the row level, VPD offers controlled access to security-relevant columns so that employees could see their own salaries, but no one else's.</p> <p>VPD is application-transparent. Security is enforced at the database layer and takes into account application-specific logic used to limit data access within the database. Both standard and custom-built applications can take advantage of the fine-grained access control, without the need to change a single line of application code.</p> <p>Within an enterprise, VPD results in a lower cost of ownership in deploying applications. Security can be built once, in the warehouse, rather than in every application that accesses data. Security is stronger, because it is enforced by the database, no matter how a user accesses data. Security cannot be bypassed by a user accessing data through an ad hoc query tool or new report writer. In an enterprise data warehouse, which often supports dozens of different applications as well as many end-user tools, Virtual private database feature is key technology.</p> <a id="TDPDW01231" name="TDPDW01231"></a> <div class="sect2"><!-- infolevel="all" infotype="General" --><a id="sthref281" name="sthref281"></a> <h3><a name="TDPDW01231|How Virtual Private Database Works"></a>How Virtual Private Database Works</h3> <p>Virtual private data<a id="sthref282" name="sthref282"></a>base is enabled by associating a security policy with a table, view, or synonym. An administrator uses the PL/SQL <code>DBMS_RLS</code> package to bind a policy function with a database object. Direct or indirect access to the object with an attached security policy causes the database to consult a function implementing the policy. The policy function returns a predicate (a <code>WHERE</code> clause) that the database appends to the user's SQL statement, thus transparently and dynamically modifying the user's data access.</p> <p>An application context enables access conditions to be based on virtually any attribute a database administrator deems significant, such as organization, subscriber number, account number, or position. For example, a warehouse of sales data can enforce access based on customer number, and whether the user is a customer, a sales representative or a marketing analyst. In this way, customers can view their order history over the Web (but only for their own orders), while sales representatives can view multiple orders, but only for their own customers, and analysts can analyze all sales from the previous two quarters.An application context acts as a secure cache of data that may be applied to a fine-grained access control policy on a particular object. Upon user login to the database, Oracle Database sets up an application context to cache information in the user's session. Information in the application context is defined by a developer based on information relevant to the particular application. For example, a reporting application that will query regional sales data can base its access control on the user's position and division. The application, in this case, could initially set up an application context for each user as he logs in and populate the context with data queried from the employees and departments tables for the user's position and division, respectively. The package implementing the VPD policy on the regional sales table references this application context to populate the user's position and division for each query. As such, an application context makes executing subqueries unnecessary, which might otherwise hinder performance.</p> </div> <!-- class="sect2" --></div> <!-- class="sect1" --> <p><a id="CDDBADHH" name="CDDBADHH"></a><a id="TDPDW0124" name="TDPDW0124"></a></p> <div class="sect1"><!-- infolevel="all" infotype="General" --> <h2><a name="TDPDW0124|Overview of Oracle Label Security"></a>Overview of Oracle Label Security</h2> <p>Oracle<a id="sthref283" name="sthref283"></a> Label Security, a security option for Oracle Database, extends the Virtual Private Database (VPD) to enforce label-based access control. Oracle Label Security is a complete, VPD-enabled application, augmenting VPD with labeled data management. It increases the ease of deploying secure data warehouses and provides row-level security out-of-the-box.</p> <p>Label-based access control lets you assign sensitivity labels to rows in a table, control access to that data based on those labels, and ensure that data is marked with the appropriate security label. For example, an organization may differentiate between company confidential information and partner information. Furthermore, there may be some confidential information that can be shared with certain key partners, and some that is only accessible by certain subsets of internal groups, such as the finance or sales divisions. The ability to manage labeled data is a great advantage for organizations to provide proper information to the appropriate people, at the proper data access level.</p> <a id="TDPDW01241" name="TDPDW01241"></a> <div class="sect2"><!-- infolevel="all" infotype="General" --><a id="sthref284" name="sthref284"></a> <h3><a name="TDPDW01241|How Oracle Label Security Works"></a>How Oracle Label Security Works</h3> <p>Oracle Label Security uses <span class="bold">policies</span>, which are collections of labels, user authorizations and security enforcement options. Once created, policies can be applied to entire application schemas or specific application tables. Oracle Label Security supports multiple policy definitions within a single data warehouse. Label definitions, user authorizations and enforcement options are defined on a per-policy basis. For example, a marketing policy might have labels such as marketing-only, manager, and senior vice president.</p> <p>Oracle Label Security mediates access to rows in database tables based on a label contained in the row, a label associated with each database session, and Oracle Label Security privileges assigned to the session. It provides access mediation on an application table after a user has been granted the standard database system and object privileges. For example, suppose a user has <code>SELECT</code> privilege on a table. If the user executes a <code>SELECT</code> statement on the table, Oracle Label Security will evaluate the selected rows and determine if the user can access them based on the privileges and access labels assigned to the user. Oracle Label Security also performs such security checks on <code>UPDATE</code>, <code>DELETE</code>, and <code>INSERT</code> statements. Labels can be applied to tables as well as to materialized views, where the materialized views increase performance and labels increase security, thus ensuring the flexibility, speed and scalability desired in data warehouse environments.</p> </div> <!-- class="sect2" --> <a id="TDPDW01242" name="TDPDW01242"></a> <div class="sect2"><!-- infolevel="all" infotype="General" --><a id="sthref285" name="sthref285"></a> <h3><a name="TDPDW01242|How Data Warehouses Benefit from Labels"></a>How Data Warehouses Benefit from Labels</h3> <p>Oracle Label Security lets you consolidate information from multiple sources into one, very large system, with the convenience and manageability and security of centralized administration. Because this security option is an application on its own, there is no need to do any PL/SQL programming. It enables consolidation, minimizes risk by enforcing security on the data itself, and provides fine-grained access security by controlling access to data down to the row level.</p> </div> <!-- class="sect2" --></div> <!-- class="sect1" --> <p><a id="CDDGEAGB" name="CDDGEAGB"></a><a id="TDPDW0125" name="TDPDW0125"></a></p> <div class="sect1"><!-- infolevel="all" infotype="General" --> <h2><a name="TDPDW0125|Overview of Fine-Grained Auditing"></a>Overview of Fine-Grained Auditing in Data Warehouses</h2> <p>Fine-grained<a id="sthref286" name="sthref286"></a> auditing allows the monitoring of data access based on content. It enables you to specify the columns and conditions that you want audit records for. Conditions can include limiting the audit to specific types of DML statements used in connection with the columns that you specify. You can also provide the name of the routine you want called when an audit event occurs. This routine can notify or alert administrators or handle errors and anomalies. An example of fine-grained auditing would be a central tax authority tracking access to tax returns to guard against employee snooping. It is insufficient to know that a specific user issued a <code>SELECT</code> statement on a particular table. What is necessary for robust security is auditing at the finer level of when a user tries to access information that is not needed to perform his normal duties, in this case, a <code>SELECT</code> statement on a column or row containing non-work related information. Fine-grained auditing offers this capability.</p> <p>Fine-grained auditing can be implemented in user applications using the <code>DBMS_FGA</code> package or by using database triggers.</p> </div> <!-- class="sect1" --> <p><a id="CDDCEBAG" name="CDDCEBAG"></a><a id="TDPDW0126" name="TDPDW0126"></a></p> <div class="sect1"><!-- infolevel="all" infotype="General" --> <h2><a name="TDPDW0126|Overview of Transparent Date Encryption in DWs"></a>Overview of Transparent Data Encryption in Data Warehouses</h2> <p>Transparent<a id="sthref287" name="sthref287"></a> data encryption enables encryption of sensitive data in database columns as the data is stored in the operating system files. It provides for secure storage and management of encryption keys in a security module external to the database. It eliminates the need to embed encryption routines in existing applications and dramatically lowers the cost and complexity of encryption. With a few simple commands, sensitive application data can be encrypted.</p> <p>Most encryption solutions require specific calls to encryption functions within the application code. This is expensive because it typically requires extensive understanding of an application as well as the ability to write and maintain software. In general, most organizations do not have the time or expertise to modify existing applications to make calls to encryption routines. Transparent data encryption addresses the encryption problem by deeply embedding encryption in Oracle Database. Note, however, that it works with direct-path loading, but not with SQL*Loader.</p> <p>Application logic performed through SQL will continue to work without modification. In other words, applications can use the same syntax to insert data into an application table and Oracle Database will automatically encrypt the data before writing the information to disk. Subsequent <code>SELECT</code> operations will have the data transparently decrypted so the application will continue to work normally.</p> </div> <!-- class="sect1" --> <!-- Start Footer --> <div class="footer"> <table class="simple oac_no_warn" summary="" cellspacing="0" cellpadding="0" width="100%"> <col width="86%" /> <col width="*" /> <tr> <td align="left"><span class="copyrightlogo">Copyright © 2007, 2009, Oracle and/or its affiliates. All rights reserved.</span><br /> <a href="./dcommon/html/cpyr.htm"><span class="copyrightlogo">Legal Notices</span></a></td> <td align="center"><a href="tdpdw_bandr.htm"><img src="./dcommon/gifs/leftnav.gif" alt="Previous" /><br /> <span class="icon">Previous</span></a> </td> </tr> </table> </div> <!-- class="footer" --> </body> </html>
Ms-Dos/Windows
Unix
Write backup
jsp File Browser version 1.2 by
www.vonloesch.de