Does 'system testing' precede 'system integration testing'?

Question

Accordin to ISTQB etc., it should. But from practice I can see it is not that clear.
An example: We have a backend (pulling data from external systems, processing...), DB, and frontend. Testers do system tests (in GUI) on the SIT environment (system integration testing).

Also to make the system testing, they need a system which is the result of the system integration testing.

So how can I say that SIT follows system testing? I think this is valid only if the other systems are not under my control (external systems)?

Answer 1

It is often risky and problematic to test the integration between your system and external systems that you do not control. For example, if your system performs financial transactions, it is daunting to test whether you perform those transactions correctly. Sometimes there are "test" versions of the external systems available for integration testing. If there are not, your options are constrained.

True story: at a previous job, one of the testers performed a boundary test while integrated with an external system of a large, international bank. This resulted in a unintentional $999,999,999 transaction. Someone eventually caught the error, and no harm was done, but it still resulted in a conference call between the bank, our company, and the United States Federal Reserve. The bank's system was ultimately at fault, but the tester was discouraged from performing that type of test in the future.

ISTQB terminology/practices aside, you should aim to perform integration tests with fake data and fake/mocked external systems first. After that, you may choose to perform live tests with live external systems, but you should perform those tests with the potential risks in mind.

Answer 2

The system test precedes the system integration test.

The goal of the system test is to test exactly one system. If this system depends on other systems, these should not be tested at the same time, but stubs should be used instead of them. The reason for this is, that it is harder to identify the root cause of observed failures, if multiple systems are tested together, compared to only one system being tested. These failures could be defects in any of the involved systems, or in the interactions between them, or it could be test data issues etc.

After the system test has successfully been performed, this system can move to the system integration test, where it tested only with other successfully tested systems, and maybe still some stubs. In the system integration test the interaction between systems is tested.

Answer 3

A scenario that comes to mind where System Testing precedes Integration testing is an expense application that I worked on, that integrated with an external banking application.

Once expenses are registered by a user in the Expense application, the Admin would then run a payroll export which would result in the user being reimbursed that amount through the external banking application. Our options on testing this integration was constrained and only tested up to the point that if the correct output files were received the test is deemed to have passed.

Answer 4

You are absolutely right. I also carry the same view. It's a tricky question to answer.

So, in your case, systems will be tested individually then integrated incrementally. So, when the testing team is executing test cases to ensure that when all the systems are combined one by one, it's is not adversely impacting the output and the results are as expected. This is system integration testing. So, once all the systems are integrated, the team tests the complete system as a whole which would be system tests in your case. So, SIT would be carried out BEFORE System testing.

Whatever the case be, whenever multiple systems are integrated, the integration testing is always followed by system testing when you test the system considering it a single entity and not as the one formed by integrating multiple systems.