If agile project procedures such as Scrum and the functionality of a relational database are mirrored, it is not surprising that many projects sooner or later run into performance or scaling problems. A core idea of the agile approach is the early use of feedback in order to implement necessary corrections as easily as possible. The risk of the overall project failing is reduced by a strategy of a " fail fast and fix quickly " approach. However, when using a relational database, this approach can mean that many details have to be rethought and redeveloped again and again.
The issue of performance is then usually seen as a problem for later administrators or even for the end customer, who will use significantly more powerful armenia telegram screening servers than actually necessary. The most common question from an agile team that has been made aware of this tension is about a pragmatic and realistic solution.
As can already be seen, in an application based on a relational database system, even small changes or extensions to a single table, view, index or procedure are enough for the query optimizer to generate new and therefore unknown query execution plans for many of the existing and tested T-SQL statements. These do not always necessarily result in a performance regression, but the risk increases significantly. This makes the isolated development or extension of individual features of an application much more difficult.
In order to ensure the functionality of the entire application, unit tests are usually used. These automatically test both the technical aspects and various error cases within the application for unexpected deviations or even errors. The issue of performance can rarely be considered.