cancel
Showing results for 
Search instead for 
Did you mean: 

Alfresco + LDAP +eDirectory

linuxos77
Champ in-the-making
Champ in-the-making
I'm trying to configure Alfresco to authenticate via our eDirectory server.  I copied the ldap-authentication-context.xml.sample to /WEB-INF/classes/alfresco/extension and made the changes shown below:


<?xml version='1.0' encoding='UTF-8'?>
<!DOCTYPE beans PUBLIC '-//SPRING//DTD BEAN//EN' 'http://www.springframework.org/dtd/spring-beans.dtd'>

<beans>

    <!– DAO that rejects changes - LDAP is read only at the moment –>

    <bean id="alfDaoImpl" class="org.springframework.transaction.interceptor.TransactionProxyFactoryBean">
        <property name="proxyInterfaces">
            <value>org.alfresco.repo.security.authentication.MutableAuthenticationDao</value>
        </property>
                <property name="transactionManager">
                        <ref bean="transactionManager" />
                </property>
                <property name="target">
                        <bean class="org.alfresco.repo.security.authentication.ntlm.NullMutableAuthenticationDao" />
                </property>
                <property name="transactionAttributes">
                <props>
                        <prop key="*">${server.transaction.mode.default}</prop>
                </props>
                </property>
        </bean>

    <!– LDAP authentication configuration –>

    <!–

    You can also use JAAS authentication for Kerberos against Active Directory or NTLM if you also require single sign on from the
    web browser. You do not have to use LDAP authentication to synchronise groups and users from an LDAP store if it supports other
    authentication routes, like Active Directory.

    –>

    <bean id="authenticationComponentImpl" class="org.alfresco.repo.security.authentication.ldap.LDAPAuthenticationComponentImpl">
        <property name="LDAPInitialDirContextFactory">
            <ref bean="ldapInitialDirContextFactory"/>
        </property>
        <property name="userNameFormat">
            <!–

            This maps between what the user types in and what is passed through to the underlying LDAP authentication.

            "%s" - the user id is passed through without modification.
            Used for LDAP authentication such as DIGEST-MD5, anything that is not "simple".

            "cn=%s,ou=London,dc=company,dc=com" - If the user types in "Joe Bloggs" the authentricate as "cn=Joe Bloggs,ou=London,dc=company,dc=com"
            Usually for simple authentication.

            –>
            <value>%s</value>
        </property>
    </bean>

    <!–

    This bean is used to support general LDAP authentication. It is also used to provide read only access to users and groups
    to pull them out of the LDAP reopsitory

    –>

    <bean id="ldapInitialDirContextFactory" class="org.alfresco.repo.security.authentication.ldap.LDAPInitialDirContextFactoryImpl">
        <property name="initialDirContextEnvironment">
            <map>
                <!– The LDAP provider –>
                <entry key="java.naming.factory.initial">
                    <value>com.sun.jndi.ldap.LdapCtxFactory</value>
                </entry>

                <!– The url to the LDAP server –>
                <!– Note you can use space separated urls - they will be tried in turn until one works –>
                <!– This could be used to authenticate against one or more ldap servers (you will not know which one ….) –>
                <entry key="java.naming.provider.url">
                    <value>ldap://LDAP_SERVER:389</value>
                </entry>

                <!– The authentication mechanism to use      –>
                <!– Some sasl authentication mechanisms may require a realm to be set –>
                <!–                java.naming.security.sasl.realm –>
                <!– The available options will depend on your LDAP provider –>                <entry key="java.naming.security.authentication">
                    <value>DIGEST-MD5</value>
                </entry>

                <!– The id of a user who can read group and user information –>
                <!– This does not go through the pattern substitution defined above and is used "as is" –>
                <entry key="java.naming.security.principal">
                    <value>cn=LOGIN,ou=its,ou=admn,o=HCNET</value>
                </entry>

                <!– The password for the user defined above –>
                <entry key="java.naming.security.credentials">
                    <value>PASSWORD</value>
                </entry>
            </map>
        </property>
    </bean>

    <!– Ldap Syncronisation support –>

    <!–

    There can be more than one stack of beans that import users or groups. For example, it may be easier
    to have a version of ldapPeopleExportSource, and associated beans, for each sub-tree of your ldap directory
    from which you want to import users. You could then limit users to be imported from two or more sub tress and ignore
    users found else where. The same applies to the import of groups.

    The defaults shown below are for OpenLDAP.

    –>


    <!– Extract user information from LDAP and transform this to XML –>

    <bean id="ldapPeopleExportSource" class="org.alfresco.repo.security.authentication.ldap.LDAPPersonExportSource">
        <!–
        The query to select objects that represent the users to import.

        For Open LDAP, using a basic schema, the following is probably what you want:
        (objectclass=inetOrgPerson)

        For Active Directory:
        (objectclass=user)
        –>
        <property name="personQuery">
            <value>(objectclass=inetOrgPerson)</value>
        </property>

        <!–
        The seach base restricts the LDAP query to a sub section of tree on the LDAP server.
        –>
        <property name="searchBase">
            <value>o=HCNET</value>
        </property>

        <!–
        The unique identifier for the user.

        THIS MUST MATCH WHAT THE USER TYPES IN AT THE LOGIN PROMPT

        For simple LDAP authentication this is likely to be "cn" or, less friendly, "distinguishedName"

        In OpenLDAP, using other authentication mechanisms "uid", but this depends on how you map
        from the id in the LDAP authentication request to search for the inetOrgPerson against which
        to authenticate.

        In Active Directory this is most likely to be "sAMAccountName"

        This property is mandatory and must appear on all users found by the query defined above.

        –>
        <property name="userIdAttributeName">
            <value>uid</value>
        </property>

        <!– Services –>
        <property name="LDAPInitialDirContextFactory">
            <ref bean="ldapInitialDirContextFactory"/>
        </property>
        <property name="personService">
            <ref bean="personService"></ref>
        </property>
        <property name="namespaceService">
            <ref bean="namespaceService"/>
        </property>

        <!–
        The path to the location of a space to use as the default home folder.
        This folder should be readable by all users, or a group to which all imported users belong.
        –>
        <property name="defaultHomeFolder">
            <value>/app:company_home</value>
        </property>

        <!–
        This property defines a mapping between attributes held on LDAP user objects and
        the properties of user objects held in the repository. The key is the QName of an attribute in
        the repository, the value is the attribute name from the user/inetOrgPerson/.. object in the
        LDAP repository.
        –>
        <property name="attributeMapping">
            <map>
                <entry key="cm:userName">
                    <!– Must match the same attribute as userIdAttributeName –>
                    <value>uid</value>
                </entry>
                <entry key="cm:firstName">
                    <!– OpenLDAP: "givenName" –>
                    <!– Active Directory: "givenName" –>
                    <value>givenName</value>
                </entry>
                <entry key="cm:lastName">
                    <!– OpenLDAP: "sn" –>
                    <!– Active Directory: "sn" –>
                    <value>sn</value>
                </entry>
                <entry key="cm:email">
                    <!– OpenLDAP: "mail" –>
                    <!– Active Directory: "???" –>
                    <value>mail</value>
                </entry>
                <entry key="cm:organizationId">
                    <!– OpenLDAP: "o" –>
                    <!– Active Directory: "???" –>
                    <value>o</value>
                </entry>
            </map>
        </property>
    </bean>

    <!– Extract group information from LDAP and transform this to XML –>

    <bean id="ldapGroupExportSource" class="org.alfresco.repo.security.authentication.ldap.LDAPGroupExportSource">
        <!–
        The query to select objects that represent the groups to import.

        For Open LDAP, using a basic schema, the following is probably what you want:
        (objectclass=groupOfNames)

        For Active Directory:
        (objectclass=group)
        –>
        <property name="groupQuery">
            <value>(objectclass=groupOfNames)</value>
        </property>

        <!–
        The seach base restricts the LDAP query to a sub section of tree on the LDAP server.
        –>
        <property name="searchBase">
            <value>dc=alfresco,dc=org</value>
        </property>

        <!–
        The unique identifier for the user. This must match the userIdAttributeName on the ldapPeopleExportSource bean above.
        –>
        <property name="userIdAttributeName">
            <value>uid</value>
        </property>

        <!–
        An attribute that is a unique identifier for each group found.
        This is also the name of the group with the current group implementation.
        This is mandatory for any groups found.

        OpenLDAP: "cn" as it is mandatory on groupOfNames
        Active Directory: "cn"

        –>
        <property name="groupIdAttributeName">
            <value>cn</value>
        </property>

        <!–
        The objectClass attribute for group members.
        For each member of a group, the distinguished name is given.
        The object is looked up by its DN. If the object is of this class it is treated as a group.
        –>
        <property name="groupType">
            <value>groupOfNames</value>
        </property>

        <!–
        The objectClass attribute for person members.
        For each member of a group, the distinguished name is given.
        The object is looked up by its DN. If the object is of this class it is treated as a person.
        –>
        <property name="personType">
            <value>inetOrgPerson</value>
        </property>
        <property name="LDAPInitialDirContextFactory">
            <ref bean="ldapInitialDirContextFactory"/>
        </property>
        <property name="namespaceService">
            <ref bean="namespaceService"/>
        </property>

        <!–
        The repeating attribute on group objects (found by query or as sub groups)
        used to define membership of the group. This is assumed to hold distinguished names of
        other groups or users/people; the above types are used to determine this.

        OpenLDAP: "member" as it is mandatory on groupOfNames
        Active Directory: "member"

        –>
        <property name="memberAttribute">
            <value>member</value>
        </property>

        <property name="authorityDAO">
            <ref bean="authorityDAO"/>
        </property>
    </bean>

    <!– Job definitions to import LDAP people and groups –>
    <!– The triggers register themselves with the scheduler –>
    <!– You may comment in the default scheduler to enable these triggers –>
    <!– If a cron base trigger is what you want seee scheduled-jobs-context.xml for examples. –>

    <!– Trigger to load poeple –>
    <!– Note you can have more than one initial (context, trigger, import job and export source) set –>
    <!– This would allow you to load people from more than one ldap store –>

    <bean id="ldapPeopleTrigger" class="org.alfresco.util.TriggerBean">
        <property name="jobDetail">
            <bean id="ldapPeopleJobDetail" class="org.springframework.scheduling.quartz.JobDetailBean">
                <property name="jobClass">
                    <value>org.alfresco.repo.importer.ImporterJob</value>
                </property>
                <property name="jobDataAsMap">
                    <map>
                        <entry key="bean">
                            <ref bean="ldapPeopleImport"/>
                        </entry>
                    </map>
                </property>
            </bean>
        </property>
        <!– Start after 30 seconds of starting the repository –>
        <property name="startDelay">
            <value>30000</value>
        </property>
        <!– Repeat every hour –>
        <property name="repeatInterval">
            <value>3600000</value>
        </property>
        <!– Commented out to disable
        <property name="scheduler">
            <ref bean="schedulerFactory" />
        </property>
        –>
    </bean>

    <bean id="ldapGroupTrigger" class="org.alfresco.util.TriggerBean">
        <property name="jobDetail">
            <bean id="ldapGroupJobDetail" class="org.springframework.scheduling.quartz.JobDetailBean">
                <property name="jobClass">
                    <value>org.alfresco.repo.importer.ImporterJob</value>
                </property>
                <property name="jobDataAsMap">
                    <map>
                        <entry key="bean">
                            <ref bean="ldapGroupImport"/>
                        </entry>
                    </map>
                </property>
            </bean>
        </property>
        <!– Start after 30 seconds of starting the repository –>
        <property name="startDelay">
            <value>30000</value>
        </property>
        <!– Repeat every hour –>
        <property name="repeatInterval">
            <value>3600000</value>
        </property>
        <!– Commented out to disable
        <property name="scheduler">
            <ref bean="schedulerFactory" />
        </property>
        –>
    </bean>

    <!– The bean that imports xml describing people –>

    <bean id="ldapPeopleImport" class="org.alfresco.repo.importer.ExportSourceImporter">
        <property name="importerService">
            <ref bean="importerComponent"/>
        </property>
        <property name="transactionService">
            <ref bean="transactionComponent"/>
        </property>
        <property name="authenticationComponent">
            <ref bean="authenticationComponent"/>
        </property>
        <property name="exportSource">
            <ref bean="ldapPeopleExportSource"/>
        </property>

        <!– The store that contains people - this should not be changed –>
        <property name="storeRef">
            <value>${spaces.store}</value>
        </property>

        <!– The location of people nodes within the store defined above - this should not be changed –>
        <property name="path">
            <value>/${system.system_container.childname}/${system.people_container.childname}</value>
        </property>

        <!– If true, clear all existing people before import, if false update/add people from the xml –>
        <property name="clearAllChildren">
            <value>false</value>
        </property>
        <property name="nodeService">
            <ref bean="nodeService"/>
        </property>
        <property name="searchService">
            <ref bean="searchService"/>
        </property>
        <property name="namespacePrefixResolver">
            <ref bean="namespaceService"/>
        </property>


        <property name="caches">
            <set>
                <ref bean="permissionsAccessCache"/>
            </set>
        </property>
    </bean>

    <!– The bean that imports xml descibing groups –>

    <bean id="ldapGroupImport" class="org.alfresco.repo.importer.ExportSourceImporter">
        <property name="importerService">
            <ref bean="importerComponent"/>
        </property>
        <property name="transactionService">
            <ref bean="transactionComponent"/>
        </property>
        <property name="authenticationComponent">
            <ref bean="authenticationComponent"/>
        </property>
        <property name="exportSource">
            <ref bean="ldapGroupExportSource"/>
        </property>
        <!– The store that contains group information - this should not be changed –>
        <property name="storeRef">
            <value>${alfresco_user_store.store}</value>
        </property>

        <!– The location of group information in the store above - this should not be changed –>
        <property name="path">
            <value>/${alfresco_user_store.system_container.childname}/${alfresco_user_store.authorities_container.childname}</value>
        </property>

        <!– If true, clear all existing groups before import, if false update/add groups from the xml –>
        <property name="clearAllChildren">
            <value>true</value>
        </property>
        <property name="nodeService">
            <ref bean="nodeService"/>
        </property>
        <property name="searchService">
            <ref bean="searchService"/>
        </property>
        <property name="namespacePrefixResolver">
            <ref bean="namespaceService"/>
        </property>

        <!– caches to clear on import of groups –>
        <property name="caches">
            <set>
                <ref bean="userToAuthorityCache"/>
                <ref bean="permissionsAccessCache"/>
            </set>
        </property>

        <!– userToAuthorityCache –>
    </bean>

</beans>

My service starts but does seem to authenticate what so ever.  Aside the the "file server is wrong authenticator" message I don't see anything wrong in the logs.  I am admittedly a beginner at Alfresco, am I missing something in my config file shown above?  Thanks
12 REPLIES 12

andy
Champ on-the-rise
Champ on-the-rise
Hi

Can you log in as one of the users in your ldap store via the UI?


To sync users and groups you need to remove the comments around


       <!– Commented out to disable
        <property name="scheduler">
            <ref bean="schedulerFactory" />
        </property>
        –>

To schedule th sync up.

Regards

Andy

linuxos77
Champ in-the-making
Champ in-the-making
I think we're getting closer…  I removed the comments around the scheduler.  Now when the service comes up it looks at out LDAP tree (I'm running a network scan to watch the traffic).  However, my logins are still failing.  The catalina.out log shows the following error:

12:42:31,507 ERROR [quartz.core.JobRunShell] Job DEFAULT.ldapPeopleJobDetail threw an unhandled Exception:
org.alfresco.repo.importer.ExportSourceImporterException: Failed to import
        at org.alfresco.repo.importer.ExportSourceImporter.doImport(ExportSourceImporter.java:206)
        at org.alfresco.repo.importer.ImporterJob.execute(ImporterJob.java:36)
        at org.quartz.core.JobRunShell.run(JobRunShell.java:191)
        at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java:516)
Caused by: org.alfresco.repo.importer.ExportSourceImporterException: User returned by user search does not have mandatory user id attribute

Is there a way to know what attribute it's tripping on?

andy
Champ on-the-rise
Champ on-the-rise
Hi

One of your users in LDAP does not have a uid attribute (as specified in the import) - this is mandatory. (1.4 can be confugured to skip such entries)
You can fix your ldap query to exclude these users or fix them up in your LDAP repository.

Regards

Andy

linuxos77
Champ in-the-making
Champ in-the-making
I'll have to check with our eDir admin and see what's going on with that.  Do I use the same LDAP.xml file to configure LDAP in 1.4?  If so I'll try that.  We're still about 6-12 months from considering deployment (if this demo works we would probably purchase Ent edition).

andy
Champ on-the-rise
Champ on-the-rise
Hi

1.3 and 1.4 config are similar. There have been a few bug fixes so I would start again with the 1.4 example file.

The query is easy to fix up ….


(objectclass=inetOrgPerson)

becomes


(& (objectclass=inetOrgPerson)(uid=*))

This means only inetOrgPersons that have a vald uid are found (or whatever attribute you use for this).

Regards

Andy

linuxos77
Champ in-the-making
Champ in-the-making
THANK YOU!!  I can now authenticate via eDirectory!!  Now for the final question, it appears looking at the ldap properties file that some attributes should automatically get pulled in (first name, last name and email address are the ones I want).  Using the example ldap file I put in out Edirectory information and it works for authentication.  However, when I log in and go to my user properties my first name entry is the only one filled in (and it's filled in with my userid, not first name).  Our Edir server is using the correct attribute names in the information is stored in LDAP, but not getting pulled into Alfresco.  This is the final step I need before I can demo this.  I really appreciate all the help thus far.

dmrnj
Champ in-the-making
Champ in-the-making
Hi

1.3 and 1.4 config are similar. There have been a few bug fixes so I would start again with the 1.4 example file.

The query is easy to fix up ….


(objectclass=inetOrgPerson)

becomes


(& (objectclass=inetOrgPerson)(uid=*))

This means only inetOrgPersons that have a vald uid are found (or whatever attribute you use for this).

Regards

Andy

I'd be careful with this. You're inputting a value into XML with an ampersand, which means you'll need to pass it as character data, or:
<![CDATA[(& (objectclass=inetOrgPerson)(uid=*))]]>

Here's my problem, though. We're using userPrincipalName to log in, so I obviously edited that attribute above, and our objectclass is user. I went ahead and modified the ldap xml file, and when i ran, this is what was returned:

ERROR [org.quartz.core.ErrorLogger] Job (DEFAULT.ldapGroupJobDetail threw an exception.
org.quartz.SchedulerException: Job threw an unhandled exception. [See nested exception: org.alfresco.repo.importer.ExportSourceImporterException: Failed to import]
   at org.quartz.core.JobRunShell.run(JobRunShell.java:202)
   at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java:516)
* Nested Exception (Underlying Cause) —————
org.alfresco.repo.importer.ExportSourceImporterException: Failed to import
   at org.alfresco.repo.importer.ExportSourceImporter.doImport(ExportSourceImporter.java:206)
   at org.alfresco.repo.importer.ImporterJob.execute(ImporterJob.java:36)
   at org.quartz.core.JobRunShell.run(JobRunShell.java:191)
   at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java:516)
Caused by: org.alfresco.repo.importer.ExportSourceImporterException: User missing user id attribute DN =CN=Adminguy,CN=Users,DC=our,DC=domain,DC=edu  att = userPrincipalName
   at org.alfresco.repo.security.authentication.ldap.LDAPGroupExportSource.buildGroupsAndRoots(LDAPGroupExportSource.java:508)
   at org.alfresco.repo.security.authentication.ldap.LDAPGroupExportSource.generateExport(LDAPGroupExportSource.java:161)
   at org.alfresco.repo.importer.ExportSourceImporter.doImport(ExportSourceImporter.java:166)
   … 3 more

Nobody else's account returns this, because everyone else has userPrincipalName defined. And, unlike before, when I didn't have this filter in place, Alfresco is still running and I can still log in. Still, why is this a problem? Shouldn't the filter just NOT be pulling adminguy up?

andy
Champ on-the-rise
Champ on-the-rise
Hi

THANK YOU!! I can now authenticate via eDirectory!! Now for the final question, it appears looking at the ldap properties file that some attributes should automatically get pulled in (first name, last name and email address are the ones I want). Using the example ldap file I put in out Edirectory information and it works for authentication. However, when I log in and go to my user properties my first name entry is the only one filled in (and it's filled in with my userid, not first name). Our Edir server is using the correct attribute names in the information is stored in LDAP, but not getting pulled into Alfresco. This is the final step I need before I can demo this. I really appreciate all the help thus far.

It sounds like your users are not getting pulled from LDAP.
It is also possible the person information is created blank at login time - it will be updated sometime later when the update from ldap runs.

Make sure that the import jobs are scheduled, by default they are not.
Turn on debug for org.alfresco.repo.security.authentication.ldap.

Regards

Andy

andy
Champ on-the-rise
Champ on-the-rise
Hi

1.4 allows better configuration of what to do whe there are errors importing people. You can carry on and just log these issues..

You will still have your old person entries from when import worked. (If you were not auto creating people, this may have stopped you loggin in before.)

If you have a presence filter for the uid attribute all I can think of is that it matched in the search but you are not allowed to see the actual value??

Regards

Andy