摘要:
A method and apparatus for preventing rogue implementations of a security-sensitive class interface are provided. With the method and apparatus, a unique identifier (UID) is created by a server process when the server process is started. Anytime the server process, i.e. a server runtime environment, instantiates a new credential object following start-up of the server process, the encrypted UID is placed into a private field within the new credential object. In addition, the UID is encrypted and stored in a private class of the server runtime environment. A verification class is provided within the server runtime environment which includes one or more methods that receive the credential object as a parameter and return true or false as to the validity of the credential object. These one or more methods determine the validity of the credential object by retrieving the encrypted UID from the private class stored in the server runtime environment, decrypting the UID and comparing it to the decrypted UID stored in the private field of the credential object. If the two UIDs match, a determination is made that the credential object was created by the server runtime environment rather than a rogue application. If the two UIDs do not match, or if there is no UID in the credential object, then a false result will be returned by the verification class.
摘要:
A system and method for tracking user security credentials in a distributed computing environment. The security credentials of an authenticated user includes not just his unique user identifier, but also a set of security attributes such as the time of authentication, the location where the user is authenticated (i.e., intranet user v. internet user), the authentication strength, and so on. The security attributes are used in access control decisions. The same user can be given different authorization if he has a different security attribute value. Security credentials may be generated either by WebSphere security code or by third party security provider code. This invention stores the user credentials in a distributed cache and provides a system and method to compute the unique key based on the dynamic security credentials for cache lookup.
摘要:
A method and apparatus for preventing rogue implementations of a security-sensitive class interface are provided. With the method and apparatus, a unique identifier (UID) is created by a server process when the server process is started. Anytime the server process, i.e. a server runtime environment, instantiates a new credential object following start-up of the server process, the encrypted UID is placed into a private field within the new credential object. In addition, the UID is encrypted and stored in a private class of the server runtime environment. A verification class is provided within the server runtime environment which includes one or more methods that receive the credential object as a parameter and return true or false as to the validity of the credential object. These one or more methods determine the validity of the credential object by retrieving the encrypted UID from the private class stored in the server runtime environment, decrypting the UID and comparing it to the decrypted UID stored in the private field of the credential object. If the two UIDs match, a determination is made that the credential object was created by the server runtime environment rather than a rogue application. If the two UIDs do not match, or if there is no UID in the credential object, then a false result will be returned by the verification class.
摘要:
Objects on application servers may be defined into classes which receive different levels of security protection, such as definition of user objects and administrative objects. Domain-wide security may be enforced on administrative objects, which user object security may be configured separately for each application server in a domain. In a CORBA architecture, IOR's for shared objects which are to be secured on a domain-wide basis, such as administrative objects, are provided with tagged components during IOR creation and exporting to a name server. Later, when the IOR is used by a client, the client invokes necessary security measures such as authentication, authorization and transport protection according to the tagged components.
摘要:
A method and apparatus for preventing rogue implementations of a security-sensitive class interface are provided. With the method and apparatus, a unique identifier (UID) is created by a server process when the server process is started. Anytime the server process, i.e. a server runtime environment, instantiates a new credential object following start-up of the server process, the encrypted UID is placed into a private field within the new credential object. In addition, the UID is encrypted and stored in a private class of the server runtime environment. A verification class is provided within the server runtime environment which includes one or more methods that receive the credential object as a parameter and return true or false as to the validity of the credential object. These one or more methods determine the validity of the credential object by retrieving the encrypted UID from the private class stored in the server runtime environment, decrypting the UID and comparing it to the decrypted UID stored in the private field of the credential object. If the two UIDs match, a determination is made that the credential object was created by the server runtime environment rather than a rogue application. If the two UIDs do not match, or if there is no UID in the credential object, then a false result will be returned by the verification class.
摘要:
A system and method for tracking user security credentials in a distributed computing environment. The security credentials of an authenticated user includes not just his unique user identifier, but also a set of security attributes such as the time of authentication, the location where the user is authenticated (i.e., intranet user v. internet user), the authentication strength, and so on. The security attributes are used in access control decisions. The same user can be given different authorization if he has a different security attribute value. Security credentials may be generated either by WebSphere security code or by third party security provider code. This invention stores the user credentials in a distributed cache and provides a system and method to compute the unique key based on the dynamic security credentials for cache lookup
摘要:
Objects on application servers may be defined into classes which receive different levels of security protection, such as definition of user objects and administrative objects. Domain-wide security may be enforced on administrative objects, which user object security may be configured separately for each application server in a domain. In a CORBA architecture, IOR's for shared objects which are to be secured on a domain-wide basis, such as administrative objects, are provided with tagged components during IOR creation and exporting to a name server. Later, when the IOR is used by a client, the client invokes necessary security measures such as authentication, authorization and transport protection according to the tagged components.
摘要:
A method and apparatus for preventing rogue implementations of a security-sensitive class interface are provided. With the method and apparatus, a unique identifier (UID) is created by a server process when the server process is started. Anytime the server process, i.e. a server runtime environment, instantiates a new credential object following start-up of the server process, the encrypted UID is placed into a private field within the new credential object. In addition, the UID is encrypted and stored in a private class of the server runtime environment. A verification class is provided within the server runtime environment which includes one or more methods that receive the credential object as a parameter and return true or false as to the validity of the credential object. These one or more methods determine the validity of the credential object by retrieving the encrypted UID from the private class stored in the server runtime environment, decrypting the UID and comparing it to the decrypted UID stored in the private field of the credential object. If the two UIDs match, a determination is made that the credential object was created by the server runtime environment rather than a rogue application. If the two UIDs do not match, or if there is no UID in the credential object, then a false result will be returned by the verification class.
摘要:
Objects on application servers are distributed to one or more application servers; a user is allowed to declare in a list which objects residing on each application server are to be protected; the list is read by an interceptor; responsive to exportation of a Common Object Request Broker Architecture (“CORBA”) compliant Interoperable Object Reference (“IOR”) for a listed object, the interceptor associates one or more application server security flags with interfaces to the listed objects by tagging components of the IOR with one or more security flags; and one or more security operations are performed by an application server according to the security flags tagged to the IOR when a client accesses an application server-stored object, the security operations including an operation besides establishing secure communications between the client process and the server-stored object.
摘要:
An extensible token framework is provided for identifying purpose and behavior of run time security objects. The framework includes a set of marker token interfaces, which extends from a default token interface. A service provider may implement one or more marker token interfaces for a Subject or a thread of execution. A service provider may also implement its own custom marker tokens to perform custom operations. The security infrastructure runtime recognizes behavior and purpose of run time security objects based on the marker or custom marker token interfaces the token implements and handles the security objects accordingly.