FLUSH AGENT SLOWNESS -PUBLISHER

@udayaku how much content are you flushing on each page activation? Is it like 1:1 or are you using anything like ACS Cache Flush rules and ACS cache flush UI? Try to reduce number of records getting flushed in each activation also Try to check the dispatcher Throughput as technically it all depends on how performing is your publisher or how fast your dispatcher reacting?

Hello @udayaku ,

The replication agents are clearly being overwhelmed however, since you mentioned it started happening only after the upgrade. 2 things to investigate,

1. Compare the number of pages that were published before and after the upgrade.

2. Set logging to debug or trace to observe if there are newer errors that were not there before the upgrade. Splunk or a similar system could help here.

thanks,

Preetpal

And,

3. Were there any changes to the AEM architecture, for example number of the AEM publish servers changed, firewall changes, number of apache servers (or similar reverse proxy), or anything else that could have changed before and after upgrade.

4. Is there a performance testing environment that you can use to replicate the scenario with a similar content volume?

thanks,

Preetpal