Increasing the performance of an LDAP repository in a federated repository configuration

Follow the steps given here to increase the performance of an LDAP repository in a federated repository configuration.

The settings available on the Performance panel are independent options that pertain specifically to an LDAP repository configured using the federated repositories functionality. These options do not affect your entire WebSphere Application Server configuration.

Tasks

1. In the administrative console, click Security > Global security.

2. Under User account repository, select Federated repositories from the Available realm definitions field and click Configure.

   To configure for a specific domain in a multiple security domain environment, click Security domains > domain_name. Under Security Attributes, expand User Realm, and click Customize for this domain . Select the Realm type as Federated repositories and then click Configure.

3. Under Related items, click Manage repositories > repository_name.

4. Under Additional properties, click Performance.

5. Optional: Select the Limit search time option and enter the maximum number of milliseconds that the Application Server can use to search through the LDAP entries.

6. Optional: Select the Limit search returns option and enter the maximum number of entries to return that match the search criteria.

7. Optional: Select the Use connection pooling option to specify whether the Application Server can store separate connections to the LDAP server for reuse.

8. Optional: Select the Enable context pool option to specify whether multiple applications can use the same connection to the LDAP server. If we select the option, specify the initial, preferred, and maximum number of entries that can use the same connection. The Enable context pool option can be enabled either in conjunction with the Use connection pool option or separately. If this option is disabled, a new connection is created for each context. We can also select the Context pool times out option and specify the number of seconds after which the entries in the context pool expire.

9. Optional: Set the Maximum size value of the context pool to zero (0).

10. Optional: Select the Cache the attributes option and specify the maximum number of search attribute entries. This option enables WebSphere Application Server to save the LDAP entries so that it can search the entries locally rather than making multiple calls to the LDAP server. Click the Cache times out option associated with the Cache the attributes option to specify the maximum number of seconds that the Application Server can save these entries. Specify the Distribution policy for the dynamic attribute cache in a clustered environment as Not shared, shared Push, or shared Push and pull. This setting is read during the adapter startup process and the cache policy is set accordingly.

11. Optional: Select the Cache the search results option and specify the maximum number of search result entries. This option enables WebSphere Application Server to save the results of a search inquiry instead of making multiple calls to the LDAP server to search and retrieve the results of that search. Click the Cache times out option associated with the Cache the search results option to specify the maximum number of seconds that the Application Server can save the results. Specify the Distribution policy for the dynamic attribute cache in a clustered environment as Not shared, shared Push, or shared Push and pull. This setting is read during the adapter startup process and the cache policy is set accordingly.

12. Optional: Create the root DataObject object locally using the com.ibm.websphere.wim.util.SDOHelper.createRootDataObject method instead of the com.ibm.websphere.wim.ServiceProvider.createRootDataObject method.

These options are available to potentially increase the performance of our federated repositories configuration. However, the any increase in performance is dependant upon your specific configuration.

Subtopics

• LDAP performance settings
  Minimize impacts to performance by adding opened connections and contexts to internally maintained pools and reusing them. Also minimize performance impacts by maintaining internal caches of retrieved data.

Related:

Connection pooling