ION Bod Performance with Infor Application Connection Point in cloud

I am testing to see if there could be an issue with Bod throughput on one of my applications. That I push more bods than can be handled.

In a Cloud multi-tenant environment with a setup of 2 Infor Application Connection Point (SQL Server) with default configuration. I am achieving on average 15 messages per second throughput between these endpoints.

Is this the expected average throughput or there is some configuration that i am missing?

Find more posts tagged with

ION Connect

Accepted answers

All comments

antonio-caria

is not good 15 messages / second?

Legacy Contributor

I have cases where I have 60k Bods to Sync, in that case I must find another way to do it, if 15 msg/s is the norm.

Legacy Contributor

@kfreimanis

Rabbit MQ connection would be perfect for your application, you can set up a RabbitMQ server in a network push the BODs to different channels and consume them from your external application. Also, you have the provision to set the maximum number of send channels.

Legacy Contributor

Hey there, @kalvis -

When it comes to throughput, the answer is somewhat of a "it depends" situation. ION is able to scale load handling many times more than what you're seeing and while there is not a hard and fast limit on an 'upper' ceiling, we commonly see other customers processing peak loads of hundreds per second.

Are you seeing the bottleneck in ION and OneView or are you realizing this measurement on BOD delivery within either the sending or receiving application? For example, a system of record might be running batch loads to generate a number of BODs until it cycles through all eligible content in the ERP. The hang-up here could be that the sending system is gating the number of BODs it's generating to balance system performance.

Likewise, an application receiving a BOD parses the message and typically triggers some business logic in the ERP to transact. It depends on the type of BOD and how much business logic complexity there is but if you're seeing something like 15 new BODs arriving per second in the functional application, this could be something to investigate further with the integrated downstream subscriber.

If you're using the I/O box connection point type, you may realize better performance gains with an IMS implementation and if this is a custom application, you may want to trace server activity to see how scale is being distributed and determine how to tune the system with either hardware or configuration settings.

antonio-caria

that case You will not need ION, neither BOD's.....

A external app will solve the problem with n threads...

Legacy Contributor

I have simple test case I set 1000 BOD’s in outbox and wait when these 1000 are in other application inbox. My average over multiple tests is around 15 msg/s. Bod size is around 1kb. All the settings are default values. Could Enterprise Connector be a bottleneck?

antonio-caria

of course the problem will be in Connector.

that's a common problem in Cloud....

Legacy Contributor

Hey Kalvis,

That's an extra piece of useful information that was left out

Since you're doing a hybrid integration, that does add a few more moving parts to consider. One is network and another is sizing and scaling. I'd recommend taking a look at the documentation around Enterprise Connector and memory increases. This should get you pointed in the right direction to see if memory is pegging during your workload processing and, if so, how to update the memory configuration.

antonio-caria

The problem is not just about the max heap usage allowed but also the maximum of Connections on each Database Pool.

Copyright © 2025 Infor. All rights reserved.