The Embedded Base format is handy for a learning tool but isn’t a good environment for serious applications. An Embedded Base ,odb has all the table data plus queries, forms, reports and a 2008 vintage HSQL database engine all compressed into a zip archive which explains why with your large tables it takes so long to close the database. I’m surprised that you haven’t experienced data loss issues from a corrupted .obd file.
The largest data set I have used has 34,000+ rows, so I can’t say for sure how well it would work with your dataset but try using a JDBC connection where the table data and a more recent version of HSQL are stored outside the .odb file then only the queries, forms and reports are compressed into the zip archive (odb). The table data is stored in a separate file structured for HSQL and not compressed.
You can download a wizard for creating a JDBC connected database from
[Wizard] Create a new ‘split’ HSQL 2.x databaset=61183
You must manually copy the tables, queries, forms and reports from the Embedded database into the JDBC connected one.
It is usually best to create the tables before copying the data, so the field type and size are correct but be aware that once created, the GUI is disabled and the table structure in a JDBC connected database can only be edited using SQL commands.
Actually. I’ve found that with a little practice and a few templates I prefer creating tables using SQL. A tutorial on that can be found at https://wiki.documentfoundation.org/images/0/02/Base_tutorial.pdf Chapter 7 has some good examples.