The December 2020 release adds enhanced visualizations for JOIN queries. Plus some performance enhancements for loading and viewing large databases. This version will be available via auto-update to all existing users and also includes the bug fix from version 2020.11.1 to fix a (rather embarrassing) bug when saving connections without a password.
Thanks to everyone for your feedback and feature requests! Check out the details below.
Download the latest version of modelDBA here. This release includes the following upgrades.
Major
Enhanced JOIN visualizations
- Previously only JOINs that followed the JOIN…ON syntax were visualized when writing queries. Now, all JOIN types will be visualized, whether it’s a comma JOIN or JOIN…WHERE. Comparisons between tables in the WHERE clause will also be detected and visualized (thank you Chris for the feature request!).
Performance Enhancements
- I’ve received reports of lagging and timeout issues when working with larger databases (>100 tables). I’ve implemented a few performance enhancements that should improve things (thank you Phil, Roman, and Losiu for the issue report!).
- Previously, tables were loaded from the database individually, and indexes and constraints were loaded by schema. That means that when loading a database there was
Tablecount + (2 * Schemacount) + 1database requests to load a single database into modelDBA. Obviously, this was incredibly inefficient, especially for remote databases 😓 . Now, databases are loaded with only 3 database requests (one for tables, constraints, and indexes). This resulted in a 10% overall speed improvement when loading a local database of 90 tables. You should see even better performance if you have more tables or are accessing a remote database server.
- Database diagrams are now plotted outside of the main application process.
- Improved diagram dragging performance.
- Reduced the number of times that database diagram elements are re-cached.
- Previously, tables were loaded from the database individually, and indexes and constraints were loaded by schema. That means that when loading a database there was
- There are still some performance issues when using the database diagram with larger databases. I’ve identified the cause, and should have it fixed in the near future.
Minor
- Labels for staged and edited tables now show up correctly in the collapsed navigation menu.
- Improved autocomplete suggestions when using aggregation functions or subqueries in the SELECT list.
- Fixed application response when dragging zoom controls or query result tabs.
Do you have a feature request for modelDBA? Found a bug 😬 ? Let me know at [email protected]. Your feedback makes modelDBA better for everyone.
You can get the latest version of modelDBA here. Or, just open up modelDBA on your machine and you’ll be prompted to upgrade.