In SugarCRM Integration Scoping Part 2, we provided an overview of what we consider Full Integrations, which includes synchronizing a large amount of data bidirectionally. In this article, we will review the three common strategies used to build a custom integration along with their pros and cons.
Logic Hooks and Schedulers
The first strategy for a SugarCRM integration is to programmatically build the integration using SugarCRM logic hooks and schedulers. If you are not familiar with Schedulers, SugarCRM has a scheduler system that runs at given intervals. This code can connect to remote systems to push and pull data.
- Leverage SugarCRM code and knowledge.
- Reusable with other integrations if built correctly.
- No additional recurring expense like with an ETL tool
- Lots of custom work.
- Puts a lot of load on SugarCRM
ETL stands for Extract, Transform, Load. ETL tools are specialized tools built to create integrations. They have all of the required underpinning to build robust integrations.
- Built for the job at hand.
- Server load is less on Sugar and more on the ETL server.
- Integrations can be built more quickly.
- Additional tools to learn and adopt.
- Another recurring expense.
Queuing technology like RabbitMQ and IBM MQ are built for high volume data transaction and synchronization. Queuing architecture offers many different features compared to an ETL tool when it comes to moving data. Just because you use a queueing architecture doesn’t mean that you forgo the need to utilize ETL tools and Schedulers.
- Great for high volume data
- Highly redundant if built correctly.
- Adds Integration complexity
- Additional cost in conjunction with ETL and customer code
If you are building a custom integration or working with a consulting firm to build a custom integration then you will need to understand the pros and cons of the solutions that are being presented. We hope that this article provided a good primer on the options that are available.