Highlighted

AEM6.4 war deployment on WebLogic managed servers

Avatar

Avatar

sayabanna

Avatar

sayabanna

sayabanna

25-04-2019

Hi,

We have a domain which has got one weblogic admin server and two managed servers and we tried the deployment of war file on both managed servers through admin console of weblogic.

Deployment was successful but AEM instance came up in one managed server and didn't come up on other.

Please let me know if this approach of deployment is supported or any other change needs to be done to get the second instance running.

Thanks

Replies

Highlighted

Avatar

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

891

Correct Answer

1.0K

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

891

Correct Answer

1.0K
Jörg_Hoh
Employee

25-04-2019

technically it's possible to deploy instances in that way. Recommended ... I would recommend to drop WL and use the jetty-based "quickstart" deployment instead 🙂

Regarding the case where one instance did not come up: You should check the logfiles, why it did not work.

Highlighted

Avatar

Avatar

sayabanna

Avatar

sayabanna

sayabanna

26-04-2019

Hi Joerg,

Thanks for the response,

I did check the log on failed server and below is the snippet.

<26-Apr-2019, 12:27:40,527 PM BST> <Error> <HTTP> <BEA-101020> <[ServletContext@1071646774[app:AEM_6.4_Quickstart module:AEM_6.4_Quickstart.war path:null spec-version:3.1]] Servlet failed with an Exception

java.lang.NullPointerException

        at weblogic.servlet.internal.ServletResponseImpl.sendContentError(ServletResponseImpl.java:727)

        at weblogic.servlet.internal.ServletResponseImpl.sendError(ServletResponseImpl.java:779)

        at weblogic.servlet.internal.ServletResponseImpl.sendError(ServletResponseImpl.java:702)

        at javax.servlet.http.HttpServletResponseWrapper.sendError(HttpServletResponseWrapper.java:158)

        at org.apache.felix.http.base.internal.dispatch.ServletResponseWrapper.sendError(ServletResponseWrapper.java:139)

        Truncated. see log file for complete stacktrace

>

<26-Apr-2019, 12:46:18,288 PM BST> <Error> <ServletContext-/AEM_6.4_Quickstart> <BEA-000000> <ERROR: Error stopping bundle. (java.lang.NullPointerException)

java.lang.NullPointerException

        at org.apache.felix.cm.impl.persistence.PersistenceManagerTracker$4.removedService(PersistenceManagerTracker.java:331)

        at org.osgi.util.tracker.ServiceTracker$Tracked.customizerRemoved(ServiceTracker.java:967)

        at org.osgi.util.tracker.ServiceTracker$Tracked.customizerRemoved(ServiceTracker.java:870)

        at org.osgi.util.tracker.AbstractTracked.untrack(AbstractTracked.java:341)

        at org.osgi.util.tracker.ServiceTracker.close(ServiceTracker.java:377)

        Truncated. see log file for complete stacktrace

>

<26-Apr-2019, 12:46:18,290 PM BST> <Error> <ServletContext-/AEM_6.4_Quickstart> <BEA-000000> <ERROR: Error stopping slinginstall:org.apache.felix.configadmin-1.9.0-B006.jar (java.lang.NullPointerException)

java.lang.NullPointerException

        at org.apache.felix.cm.impl.persistence.PersistenceManagerTracker$4.removedService(PersistenceManagerTracker.java:331)

        at org.osgi.util.tracker.ServiceTracker$Tracked.customizerRemoved(ServiceTracker.java:967)

        at org.osgi.util.tracker.ServiceTracker$Tracked.customizerRemoved(ServiceTracker.java:870)

        at org.osgi.util.tracker.AbstractTracked.untrack(AbstractTracked.java:341)

        at org.osgi.util.tracker.ServiceTracker.close(ServiceTracker.java:377)

        Truncated. see log file for complete stacktrace

What we have observed is, There is single crx-quickstart folder for both the managed servers.

is this behaviour allowed as sometimes due file system locking two apps cannot access it at same time.

Thanks

Highlighted

Avatar

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

891

Correct Answer

1.0K

Avatar

Jörg_Hoh

Employee

Total Posts

3.0K

Likes

891

Correct Answer

1.0K
Jörg_Hoh
Employee

26-04-2019

Each managed server should have its own crx-quickstart folder. And then each server has it's own repository. Be aware that from an AEM point of view this is not a cluster!

Regarding your issue: It's hard to guess what was wrong. But if they both share the same crx-quickstart directory, that would be the first thing I would try to fix: each instance should have it's own crx-quickstart directory!

Jörg

Highlighted

Avatar

Avatar

Albin_Issac

MVP

Avatar

Albin_Issac

MVP

Albin_Issac
MVP

13-05-2019

Refer the following video for more details on setting up AEM Author and Publisher in weblogic

Regards

Albin I

www.albinsblog.com

Highlighted

Avatar

Avatar

Albin_Issac

MVP

Avatar

Albin_Issac

MVP

Albin_Issac
MVP

13-05-2019

Highlighted

Avatar

Avatar

sayabanna

Avatar

sayabanna

sayabanna

14-05-2019

Hi Albin,

Thanks for the response.

We are trying to run two author instances in both managed servers instead of using one for author and another for publish as mentioned in video.

Thanks,
Sayabanna

Highlighted

Avatar

Avatar

berliant

Employee

Avatar

berliant

Employee

berliant
Employee

14-05-2019

You need to deploy a war file to each WL managed servers separately, as such AEM author will have it'own crx-quickstart folder.

What is the use case of having two authors? Keep in mind, that despite the fact that AEM is packaged as a war file, it does not support J2EE clustering. TarMK is not horizontally scalable. If you are looking for AEM author's cluster, the only option is MongoMK.

Highlighted

Avatar

Avatar

Albin_Issac

MVP

Avatar

Albin_Issac

MVP

Albin_Issac
MVP

14-05-2019

the same steps can be followed - dont change the runmode for second instance and the remaining steps are going to be same.

Regards

Albin I