"Root node is not accessible" error when Test Connection is triggered in default replication agent on Author

Avatar

Avatar

ak481808

Avatar

ak481808

ak481808

16-02-2018

Hello Everyone,

We upgraded from AEM 6.0 to AEM 6.3 (Not sure if this is an issue related to AEM 6.3)

The default replication agent is not working as expected. When Test Connection is done, it throws below error in the logs:

16.02.2018 17:55:24.549 *ERROR* [10.150.44.28 [1518783924487] GET /etc/replication/agents.author/publish.test.html HTTP/1.1] com.day.cq.cq-audit [com.day.cq.audit.impl.DefaultAuditLog(1301)] The activate method has thrown an exception (javax.jcr.AccessDeniedException: Root node is not accessible.)

javax.jcr.AccessDeniedException: Root node is not accessible.

        at org.apache.jackrabbit.oak.jcr.session.SessionImpl$4.perform(SessionImpl.java:294)

        at org.apache.jackrabbit.oak.jcr.session.SessionImpl$4.perform(SessionImpl.java:288)

        at org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:208)

        at org.apache.jackrabbit.oak.jcr.session.SessionImpl.getRootNode(SessionImpl.java:288)

        at com.adobe.granite.repository.impl.CRX3SessionImpl.getRootNode(CRX3SessionImpl.java:138)

After this, it keeps on throwing below error -

16.02.2018 13:14:15.653 *ERROR* [FelixDispatchQueue] com.day.cq.wcm.cq-msm-core FrameworkEvent ERROR (org.osgi.framework.ServiceException: Service factory returned null. (Component: com.day.cq.wcm.msm.impl.RolloutExceptionHandlerImpl (1734)))

org.osgi.framework.ServiceException: Service factory returned null. (Component: com.day.cq.wcm.msm.impl.RolloutExceptionHandlerImpl (1734))

        at org.apache.felix.framework.ServiceRegistrationImpl.getFactoryUnchecked(ServiceRegistrationImpl.java:380)

I have tried troubleshooting steps -

1. Did a ping from author to publish VM, Ping is fine.

2. Did telnet from author to publish VM, that is also fine.

3. Checked all access in publish server for the admin user, all access is as expected)

4. The replication queue is clear.

5. However, when I do a wget from author to this URL I get this -

wget http://<publishURL>:4503/bin/receive

--2018-02-16 18:16:17--  http://<publishURL>:4503/bin/receive

Connecting to <publishURL>:4503... connected.

HTTP request sent, awaiting response... 401 Unauthorized

Authorization failed.

Could this be the problem? Please let me know if someone has faced the same sort of issue.

Accepted Solutions (1)

Accepted Solutions (1)

Avatar

Avatar

Kunwar

Employee

Avatar

Kunwar

Employee

Kunwar
Employee

01-06-2018

No it's not a known issues. Something is wrong with the repository for sure, deeper analysis on the logs would he. I'd suggest to open a daycare ticket for this.

Answers (6)

Answers (6)

Avatar

Avatar

jaskiratb1369

Avatar

jaskiratb1369

jaskiratb1369

19-12-2018

some where in code login user is trying to access root node

for eg: Node root = session.getRootNode();

To check login in  as http://localhost:4502/useradmin 

relative path as : /useradmin

below is the example of without root node access permission : see the first line in the folder path no root access

withoutrootnode.JPG

for this logged in user provide at least read access permission as below :

withrootnode.JPG

Avatar

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

991

Correct Reply

1.0K

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

991

Correct Reply

1.0K
Jörg_Hoh
Employee

04-06-2018

Can you post the full stacktraces?

Jörg

Avatar

Avatar

anupamp17770278

Avatar

anupamp17770278

anupamp17770278

04-06-2018

For me these errors keep popping up even if I add a new asset or delete an existing. So any action on a node from DAM triggers this.

Avatar

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

991

Correct Reply

1.0K

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

991

Correct Reply

1.0K
Jörg_Hoh
Employee

03-06-2018

Looks like a permission issue. Can you validate the permissions of the service users "replication service" to check if you have adapted them? Compare the permissions of your migrated instance to a fresh installation of AEM 6.3.

Jörg

Avatar

Avatar

smacdonald2008

Total Posts

12.7K

Likes

1.4K

Correct Reply

2.3K

Avatar

smacdonald2008

Total Posts

12.7K

Likes

1.4K

Correct Reply

2.3K
smacdonald2008

01-06-2018

Kunwar​ - is this a known issue. This does not happen on fresh installs - we have tested AEM Rep on 6.3/6.4 with no issues. It appears to be some sort of upgrade issue.

Avatar

Avatar

anupamp17770278

Avatar

anupamp17770278

anupamp17770278

01-06-2018

I got the same issue while triggering a custom WF. I suppose it's issue with Launcher unable to get the node but not sure. Let me know if you were able to solve this and how.