Relational and non-relational databases
Posted: Tue Apr 22, 2025 4:19 am
Databases consist of structured data storage systems. In a relational database we can define explicit relationships between data, bringing some useful information through these relationships, in addition to automating operations on related data in a unified manner.
Let's look at an example:
In a database, we have a table called “stock” and another called “sales.” In a spreadsheet approach, the user would have to update two spreadsheets for each sale. The sales spreadsheet would add information about a sale, and then the stock spreadsheet would be updated to decrease the number of items sold.
If these operations are done ecuador mobile database manually, this operation is subject to errors that can cause inconsistencies in the data. The employee may not act quickly, forget details and cause problems in stock control, such as selling a product that is no longer in stock.
In a relational database, these two tables are related in a logical way where whenever there is a sale of a certain item, in the “stock” table, its quantity is automatically decreased, and in the next sale that item will have less volume in stock.
In a non-relational database we can have independent tables with distinct, unconnected data , allowing the integrity and independence of unrelated data.
Legacy systems
In many companies, especially in the banking sector, there are old systems that use outdated technologies that have essential business functions. Reimplementing such systems using new technologies becomes a problem due to the high demand for services offered by legacy systems and their complexity. A new implementation of a legacy system can harm the availability of some service as well as the loss of data in the process.
It is then necessary to integrate these systems through the existing interface, which can be through command terminals and, for not so old systems, through web interfaces. Unfortunately, not all programming languages have the necessary resources to deal with these scenarios, but most modern languages have means of interaction that cover this need. Another characteristic of legacy systems is the lack of structure in their data, another point that modern programming languages attack efficiently.
What determines good data integration?
Good integration must take into account time and data availability factors . The source from which the data is being extracted must be available at all times.
Good integration is also characterized by saving resources when obtaining that data. If your programming language does not have the necessary resources to obtain that data with quality, consider a change, and do not worry about this change because different systems have different integration requirements.
What determines poor integration?
In addition to the characteristics that are antagonistic to the previous topic, an unfeasible integration is mainly marked by the dynamicity of the location of the data source . A system that does not have a fixed point from which data is extracted is completely unfeasible.
A system to be well integrated must have a fixed integration interface, whether via API or CSV, this interface must be immutable. Otherwise, this integration will be marked as unfeasible and cannot be executed.
Let's look at an example:
In a database, we have a table called “stock” and another called “sales.” In a spreadsheet approach, the user would have to update two spreadsheets for each sale. The sales spreadsheet would add information about a sale, and then the stock spreadsheet would be updated to decrease the number of items sold.
If these operations are done ecuador mobile database manually, this operation is subject to errors that can cause inconsistencies in the data. The employee may not act quickly, forget details and cause problems in stock control, such as selling a product that is no longer in stock.
In a relational database, these two tables are related in a logical way where whenever there is a sale of a certain item, in the “stock” table, its quantity is automatically decreased, and in the next sale that item will have less volume in stock.
In a non-relational database we can have independent tables with distinct, unconnected data , allowing the integrity and independence of unrelated data.
Legacy systems
In many companies, especially in the banking sector, there are old systems that use outdated technologies that have essential business functions. Reimplementing such systems using new technologies becomes a problem due to the high demand for services offered by legacy systems and their complexity. A new implementation of a legacy system can harm the availability of some service as well as the loss of data in the process.
It is then necessary to integrate these systems through the existing interface, which can be through command terminals and, for not so old systems, through web interfaces. Unfortunately, not all programming languages have the necessary resources to deal with these scenarios, but most modern languages have means of interaction that cover this need. Another characteristic of legacy systems is the lack of structure in their data, another point that modern programming languages attack efficiently.
What determines good data integration?
Good integration must take into account time and data availability factors . The source from which the data is being extracted must be available at all times.
Good integration is also characterized by saving resources when obtaining that data. If your programming language does not have the necessary resources to obtain that data with quality, consider a change, and do not worry about this change because different systems have different integration requirements.
What determines poor integration?
In addition to the characteristics that are antagonistic to the previous topic, an unfeasible integration is mainly marked by the dynamicity of the location of the data source . A system that does not have a fixed point from which data is extracted is completely unfeasible.
A system to be well integrated must have a fixed integration interface, whether via API or CSV, this interface must be immutable. Otherwise, this integration will be marked as unfeasible and cannot be executed.