In our previous blogs, we covered the pitfalls of message validation. What happens when...
Message validation should not be dependent on a specific environment
In our previous blog, we talked about why having a specific person in charge of all testing in an organisation is alarming. Today, our topic is using a copy of the production system for message validation as a part of deployments. It may seem like an efficient solution, but is it, after all?
In this model, message validation requires an operational EDI connection, which may take days or weeks to set up, depending on the case. When the delay is repeated with every deployment, your business will suffer a tangible loss of profits.
In production systems, the meaning of message validation is to detect critical errors. In other words, this kind of validation does not recognize non-critical errors that nevertheless affect the functionality and automation of data transmission. Furthermore, feedback on errors from production systems is invariably limited, technical, and difficult to align.
Production systems will usually produce feedback asynchronously, i.e. with some delay. Combined with unclear error notifications, this makes iterative error correction numbingly difficult. Unclear feedback on errors will also trigger several support requests that first or second-level support does not have the tools to resolve. Thus, work requests flow to the expert level, which means more time and resources are spent on correcting errors.
How can we improve the current situation?
The same tests can be run easily as an iterative process without bottlenecks when message validation is carried out as a separate function, independently of the testing channel and environment. Furthermore, when validation is separated from the application code, both the validation process and the resulting feedback on errors can be updated in a more agile fashion.
In the context of deploying EDI connections, message validation should also test non-critical requirements affecting on efficiency and reliability. Maintaining parallel sets of test assertions can be avoided by using conditional requirements, which means the same test assertions can be used for both deployments and production.
Would it not make life easier if problems could be sorted out quickly and without intermediaries by employing a testing solution that offers instant feedback on errors in plain language? And if customer support could pass on feedback on errors easily and without cut-and-paste routines? Would it also not be a better use of your resources if you could properly focus on the most important function of the production environment-based testing environment, i.e. ensuring the functionality of your processes before they are applied to production?
Truugo offers the best possible solution for enabling easy and independent message validation in different contexts. If you require any assistance in developing your operations, we will be glad to help. Further information on our website at: https://www.truugo.com/
Leave your comments below to let us know what kind of problems you have encountered in this area!