Replication gets stuck frequently | Can we detect this in time?

[Shallu_Rohilla]

Hi,

After every few days we see issues with replication getting stuck with no items in queue. We see exceptions like:

com.day.cq.replication.ReplicationException: Repository error during node import: Unable to update a stale item: item.save()

However when we restarted our custom auto replicate service which checks if pages are from specific path and auto replicates them via replicator service(it basically listens for resource change), we were able to resume. 

How can a CQ replication problem/outage be detected / alerted at time of the problem starting, rather than waiting on users noticing stuff missing / not appearing?Also what is the cause of this frequent replication getting stuck? Any help would be appreciated.

Note: We are using CQ 5.6.

Can someone pls help?

Regards,

Shallu Rohilla

[Shallu_Rohilla]

Can someone please provide with some pointers?

Regards,

Shallu

[edubey]

1. Are you replicating pages via code ?

2. How frequently its happening?

Here are some help you refer:

https://helpx.adobe.com/experience-manager/kb/experience-manager-replication-faqs.html

https://helpx.adobe.com/experience-manager/kb/replication-stuck.html

Thanks

[Shallu_Rohilla]

Yes, we are auto replicating our pages as soon as they are modified via a listener.

[Opkar_Gill]

You can monitor the Replication agents using JMX Beans[0] if you do this frequently it can alert your admins to issues with replication. There is a QueueBlocked parameter you can check.

Regards,

Opkar

https://docs.adobe.com/docs/en/cq/5-6-1/core/administering/jmx-console.html

[albini43371541]

You can try this - http://www.albinsblog.com/2017/02/how-to-monitor-replication-queuesAdobeCQ5AEM.html#.WLuB71UrLIU