Configuration manager in enterprise computing system
A system and method are described for managing a plurality of application servers. In one embodiment, the application servers are organized into groups referred to as “instances.” Each instance includes a group of redundant application servers and a dispatcher for distributing service requests to each of the application servers. A group of instances may be organized as a “cluster.” Each server includes a configuration manager to facilitate changes to configuration information within the cluster. The configuration manager may include a configuration cache and monitor its consistency with respect to other instances within the cluster.
1. A method comprising:
storing a configuration for a distributed environment, which includes a first node, in a central storage of the distributed environment; and
updating a portion of the configuration by the first node in the distributed environment, the first node separate from the central storage, wherein updating comprises:
acquiring a lock for the portion of the configuration by the first node in the distributed environment;
modifying the portion of the configuration:
invalidating a representation of the portion of the configuration in the distributed environment;
releasing the lock;
updating a database at the central storage to reflect modifications of the portion of the configuration; and
blocking reads of the configuration from the database during the updating.
2. The method of claim 1 wherein updating further comprises:
notifying nodes in the distributed environment that the configuration has been updated.
3. The method of claim 1 wherein the lock is cluster wide.
4. The method of claim 1 wherein updating further comprises:
writing changes to a shared database at the central storage.
5. The method of claim 1 wherein modifying comprises:
changing a configuration object in a branch of a tree structure.
6. The method of claim 1 wherein invalidating comprises:
sending a cache invalidation event to another node in the cluster.
7. The method of claim 1 wherein invalidating comprises:
sending a message to a plurality of Java 2 Enterprise Edition (J2EE) nodes.
8. The method of claim 1 wherein updating further comprises:
notifying a registered listener at a second node that the configuration has been changed, the second node separate from the first node and the central storage.
9. A system comprising:
a plurality of nodes each having an instance of a configuration manager to maintain consistent storage of a configuration across the nodes without passing configuration modifications between the nodes, a node in the plurality of nodes to update a portion of the configuration by:
acquiring a lock for the portion of the configuration by the node;
modifying the portion of the configuration;
invalidating a representation of the portion of the configuration;
releasing the lock;
updating a shared relational database to reflect modifications of the portion of the configuration; and
blocking reads of the configuration during update;
a locking server shared by the plurality of nodes to coordinate access to the configuration; and
a database management system to provide an interface with the shared relational database, the shared relational database to store the configuration.
10. The system of claim 9 wherein the configuration manager comprises:
a configuration cache; and
a configuration handler.
11. The system of claim 10 wherein the configuration manager further comprises:
a persistency handler.
12. The system of claim 9 further comprising:
a configuration handler to permit access to and modification of the configuration.
13. The system of claim 9 wherein the configuration comprises:
a plurality of persistent objects holding information about a Java 2 enterprise edition cluster.
14. The system of claim 13 wherein some of the persistent objects are cacheable.
15. The system of claim 9 wherein the configuration manager comprises:
a change event listener to notify registered components of configuration change events.
16. A computer readable storage media containing executable computer program instructions which when executed cause a digital processing system to perform a method comprising:
storing a configuration for a distributed environment, which includes a first node, in a central storage of the distributed environment; and
updating a portion of the configuration by the first node in the distributed environment, the first node separate from the central storage, wherein updating comprises:
acquiring a lock for the portion of the configuration by the first node in the distributed environment;
modifying the portion of the configuration;
invalidating a representation of the portion of the configuration in the distributed environment;
releasing the lock;
updating a database at the central storage to reflect modifications of the portion of the configuration; and
blocking reads of the configuration from the database during the updating.
17. The computer readable storage media of claim 16 containing executable computer program instructions which when executed cause a digital processing system to perform the method wherein updating comprises:
notifying nodes in the distributed environment that the configuration has been updated.
18. The computer readable storage media of claim 16 containing executable computer program instructions which when executed cause a digital processing system to perform the method wherein updating further comprises:
changing the configuration locally;
writing the changes to a shared database at the central storage; and
committing the changes.
19. The computer readable storage media of claim 16 containing executable computer program instructions which when executed cause a digital processing system to perform the method wherein invalidating comprises:
sending a cache invalidation event to another node in the cluster.
20. The computer readable storage media of claim 16 containing executable computer program instructions which when executed cause a digital processing system to perform the method wherein updating comprises:
notifying a registered listener at a second node that the configuration has been changed, the second node separate from the first node and the central storage.
21. A system comprising:
means for maintaining consistent storage of configuration information in a distributed environment;
means for controlling write access to the configuration information by nodes of the distributed environment;
means for interfacing with a relational database system to provide persistent storage of the configuration information; and
means for updating a portion of the configuration information by the nodes of the distributed environment, the nodes separate from the relational database system, wherein updating comprises:
acquiring a lock for the portion of the configuration information by the nodes of the distributed environment;
modifying the portion of the configuration information;
invalidating a representation of the portion of the configuration information in the distributed environment;
releasing the lock;
updating the relational database system to reflect modifications of the portion of the configuration information; and
blocking reads of the configuration information during the updating.
22. The system of claim 21 wherein the configuration information comprises:
a plurality of persistent objects holding information about a Java 2 Enterprise Edition cluster.
23. The system of claim 21 wherein the means for maintaining comprises:
a configuration cache resident in each node of the distributed environment; and a configuration handler resident in each node of the distributed environment.