Low Throughput in Oracle to Azure Event Hub Replication -Seeking Optimization Advice

[Chirag_]

Hi Team.

We are currently using Oracle 19.21 as a source and Microsoft Azure Event Hub as a target with the default task settings. However, we have observed a relatively low throughput of around 1500 forfull load and CDC Task.We are seeking guidance on any adjustments we can make to the settings to improve the replication process's speed.

Regards,
Chirag Patil.

[Heinvandenheuvel]

Not much to go on here. If you cannot provide more, more useful, details then you should look at Qlik Professional Services help you on your way.

All task tuning settings default.

How wide are the tables? Many large lobs?

It's probably best to focus on just full-load first.put source_unload and target_load logging to debug and evaluate the log. 

Next CDC and enable Performance trace logging to understand source latency vs target latency. 

Maybe share a task export json and a task log?

Hein. 

[Jon_Donker]

Is the source and destination close to each other?  We had performance problems going from an Onprem database to a cloud database; just because of the latency of the round trip.

Try (ontop of what Heinvandenheuvel mentioned)

1. Try replicating to a NULL target; this will confirm or rule out whether it is a source or a target latency

2. Are there any network components in the way that might be adding to the latency?  Like QOS devices or IPS that might be adding to.  Is the network bandwidth wide enough to your target to handle the volume?

[SushilKumar]

hello @Chirag_ 

Request you to upload the diagnostic package/log/DDL, and attached the relevant info/files to for further analysis as we would like to know the cause of latency and Diagnostics package is the way to go.  

Please refer to Replicate https://community.qlik.com/t5/Knowledge/How-to-collect-Diagnostics-Package-from-Qlik-Replicate/ta-p/...

Regards,

Sushil Kumar

[Chirag_]

Hi Team,
Thank you for the replies,
It seemed to be a issue with the network which is been resolved,
thank you for the support.
Regards, 

Chirag

[john_wang]

Glad to hear that. Thank you so much for your great support ! @Chirag_