EAI : good practices

2019-03-11T14:45:39+02:002019-03-11|Good Practices|

ERP, BI systems, CRM, social networks…The Information System (IS) of a company is certainly less confusing that the inventory of Jacques Prévert. But it still is not a cohesive reference. Enterprise Application Integration (EAI) is a way of designing software architecture centred on inter-application exchanges. The principle objectives are often the rationalization of commercial operations and the improvement of quality by eliminating redundant data input. Apply the good practices…

The Information System of a company is a delicate assembly of independent applications. It is composed of multiple infrastructures and applications which provide diverse services on which the company depends to carry out its activities.

But this diversity actually puts these activities at a disadvantage. The time spent to develop inter-applicative connectors is counter-productive. According to the Gartner group, 75% of companies have more than six database systems. Their management can bring about three main pitfalls which can affect global performance:

  • Loss of extra employee time and effort to transfer the data relative to processes from one application to the next
  • Missing or uncertain data, manual control of data coherence, frequent corrections
  • Exportation practice and importation of information, or even manual data input already included in one system to other systems

Entreprise Application Integration

Practices that penalize the efficiency of teams. It’s the reason for which companies turn towards Enterprise Application Integration. This platform allows existing applications of a company to be brought together around an engine of communal application integration (middleware). EAI reforms the way of thinking of the IS by placing the principles of communication at the heart of both the conception and the architecture, interoperability and service exchange.

One application can, therefore be, in turn, consumer and service provider. Applications communicate with the architecture by means of connectors. The latter serve as an interface between EAI and applications. They examine application events and transmit associated data to EAI (or provide data to the application originating from this).

To summarize, its functionalities are triple:

  • Connection to applications
  • Conversion of information in a communal language
  • Transport of information, from the issuing application to the receiving application

On paper, EAI seems perfect. But according to a report from ebizQ, nearly 70% of EAI projects fail for management reasons. It is vital to rely on different ‘’good practices’’.

Incorporating a Skill Center

The control of EAI can become very difficult. This can have a negative impact of the levels of service at the company scale and beyond. It is necessary to put in place a collection of disciplines, tools and procedures to deal with questions such as management of the capacity and harmonisation of load balancing, security, change management and monitoring. To guarantee a harmonious and well-structured system, skill centres should be implemented.

Establishing complete documentation

Once integrated, it may be that you find a lot of information has no more importance. However, this information can turn out to be very useful and crucial when the EAI system is developed. For example, operational demands can differ when a company passes the production point. Consequently, it is recommended to keep all sorts of documentation and files (definitions, interfaces, flows, structures, collection statistics…).

Adding new services with precaution

Development and maintenance are the two most difficult steps of application integration. They demand extra effort and increase the complexity of IS. It is therefore recommended to integrate new applications or operating systems for when there is a critical demand.

Refraining from implementing useless features

This advice completes the former. The cost and complexity of EAI increases when useless features are implemented. It is fundamental to make the integration project ready to enter production in a realistic lead time. This goal will contribute in generating a positive return on investment (ROI).

There should also be a reliance on aspects of iPaaS (Integration Platform as a Service). Integration in the Cloud is one of the principle challenges by which companies are confronted. In order to respond to the growing need of secure and reliable integration solutions in the Cloud, several providers offer iPaas services.

iPaas and « self-service »

In simple terms, iPaaS is a platform that allows the construction and deployment of integration in the cloud and between the cloud and the company. With iPaaS, users can develop integration flows which connect applications residing in the cloud or on site, then deploying them without installing or managing material or middleware.

In conclusion, EAI are rich in features and efficient for large-scale projects and demanding environments. They allow the management of modular system integration by dealing with integration as a system task. This allows benefiting from a more flexible architecture where new pieces can be added and removed according to its needs.

Furthermore, besides data integration, a modern EAI can also include features such as network administration, security, acceleration and upgradability. But the combination of technologies and the nature of EAI solutions distributed sometimes complicates the deployment, monitoring and troubleshooting of tasks. Hence the necessity of combining different skills. In several cases, this expertise doesn’t exist within IT services or are distributed among several different people. Calling on specialists or dedicated platforms turns out to be vital.



2019-03-11T14:45:39+02:002019-03-11|Good Practices|
the digital agitator
our newsletter
media library
our ressources
the digital agitator
our blog