| > Show on single page > Show on multiple pages |
The amount of memory allocated to Squore Server has an impact on the overall performance of Squore. The following chart shows the influence of available RAM on the time needed to import a large-size project into the Squore database. This type of processing is normally the most time consuming in a typical Squore usage.
The test was performed with a project containing:
1,400,000 lines of code,
2,500 source code files,
21,000 functions,
1,100,000 base measures.
Running on a Windows Server 2003 64 bits machine, Squore is able to handle the project with 2GB of RAM. It could also manage this project with less memory, but at the expense of performance. It is however recommended to allow at least 4 GB of RAM to Squore Server. 8 GB may be required for large projects.
Using projects of increasing size, the following chart shows how all stages of the Squore project creation evolve linearly. These stages are respectively:
Pre-processing: creating and managing the project data.
Computation: running the analysis model on the project data.
DB update: importing the artefacts, the base and derived measures values into the Squore database.
These measurements are associated to a new project, which means that all objects are inserted into the database. In the case of a project being updated (i.e. a new version of an existing project), only the new data is inserted into the database, resulting in a much reduced DB update time.
