In today’s post we are going to explain the dissociation process, but before that, let’s recall how we got here and what is coming now.
We can see that we are reaching the lasts milestones of the project, also to the most pleasing ones. From the previous summary, elements from six to nine identify the actual topic.
A great amount of the time spent in this process are tests and debug executions, they are essential now. Let’s be honest, we are not perfect either, we know that the first run will probably throw some errors, among them, dissociators not configured correctly, mistakes at icaria’s configuration, etc. It’s time to fix them.
When we see evidences that the installation is close to be finished, we start with complete executions with persistence, that will give a beta result that the client’s developers will use in their daily work to identify new requests or to check for incompatibilities with their systems.
At this stage in the development we analyze some other requirements, such us, limitations in the run-time window. To tackle this problem, we evaluate the tables that takes longer, the largest, whether their primary keys are well defined or not, modify connection parameters or apply some of the solutions that we have to develop in the last months and that let us do dissociations with thousands of millions of records under a very tight schedule.
Another point to emphasize about the need for these first controlled executions is the fact that Icaria Mirage has high adaptation capacities. Recently there has been a need to implement a connection with an unfamiliar database that has required changes that requires attention during the first iterations and continuous development throughout the project, as an example, we had to adapt the impossibility of using batch updates or the condition of using a very small number of simultaneous connections to the database.
As we already anticipated, we are in the final stretch of the project, where the other team is already very familiarized with Icaria Mirage, the results are available and the production phase is very close. However, we have not finished yet, in the next posts we will explain the possible dissociation ways for the databases and we will illustrate the release of the production software.