A large number of objects

Sorry for the Google translation.
ScadaBR I'm testing on a large number of objects (~ 1000). Total 7000 channels. Real objects.
The result revealed some highlights:
1) this week database size increased to 350 MB. I see the following reasons:
   * Pointvalue field is of type Double, although necessary and sufficient Float;
   * Data type is better to store the properties of a data point, not in the dataType
   * Field size for the time stamp is redundant. There is no need to keep such a large number.
2) Interface great retarding, downloading all the points in the watch list. It is necessary to break the list into pages.
3) When you import a project with the national character set is lost. Although the import via the web is all right.
4) Help for the Russian language is not available. You need to move the files from the ru (there is a subdirectory ru). By the way still need to re-encode files in cp1251, or they are incorrect.
5) It is awkward that the links to the screens depending on their index in the database, and not on their xid, as would be correct. This is especially noticeable if the project is automatically generated.

Hello Serge!

ScadaBR ships with Derby, a simple DB mostly use for configuration, not really suitable for SCADA projects but it's standalone =]. Suggestions - MySQL, MSSQL, PostgreSQL or Oracle!

For the interface issues, there are some ways to tackle this, by using hierarchy, watchLists and pre-processing for historic graphics. We use the same interface in projects with over 20k points, so, it's feasable. Some tricks had to be learned.

Thanks for the feedback, we do believe this code can improve! We think much has to be changed for the next full release (AKA 2.0). I'll try to point you out some:

  • Better pointValue DB design, maybe TSDB;
  • Template creation;
  • Refactoring in the interfaces (specially views);
  • New graphics engine (I'd rater use a client-side engine than a server side);

The log goes on... If you wanna help, we'd appreciate have a hand =]