A cloud version of Glue will be available soon for small businesses, and customers using cloud storage, including Azure and AWS. Glue will always run faster though if you install it alongside your database.
Glue takes a model of your database, and uses this to construct relevant queries to find all the data for your chosen report template (or automatic report). You can build one model for your whole database(s) and Glue will glue it all together.
Glue is self hosted for many reasons. It allows you to place Glue right next to your data for optimum performance, and it gives you absolute certainty that your data remains in your control. In other words, you can use Glue onsite in most cases without updating your GDPR policy*.
All your models, routes and reports are stored in your installation for maximum security.
* You may need permission from your customers to process their data
Glue receives updates from Fjord. When downloading Glue, you'll be given a product key to enter during installation. This key helps us deliver the correct updates for your version of Glue.
Updates can be installed automatically (using a cron job) or manually
Fjord collects minimal usage data. This is data created when you use Glue, and does not contain anything about the data you store. Your models remain private to your installation (unless you choose to share them with our support team).
We use usage data to improve Glue by tracing which features are used most, and the time they take to run. Run times are normalised against your installation's private metrics before dispatch, so we can't reconstruct anything about your installation.
Glue collects data for internal use to help optimise the router and predict the response time to any given query. This helps Glue run efficiently, and never leaves your installation.