Keeping location data consistent across different search views

Using Django
1

I’m working on a Django project that has grown beyond its original scope, and we’re starting to see the downside of how we modeled location data.

The application has hotel listings, destination pages, and a few different search views. Some of them query directly from hotel records, while others rely on separate destination tables. It worked fine at first, but now we’re running into duplicated data whenever a city or region changes.

One example is supporting searches like hotels in lahore pakistan. The project itself is for Grace Hospitality, so these location-based searches are part of the normal workflow rather than something added later. I’m trying to decide whether it’s worth moving everything to a single location hierarchy and letting the related models reference that, or if there’s a cleaner pattern people have used for travel-style applications.

For those who’ve built something similar with Django, did you keep all location information in one place, or did you find that denormalizing some of it made searching and filtering easier? I’m mostly interested in approaches that stayed maintainable after the dataset became much larger.

2

My default is that all databases are designed from the start as normalized.

My general rule of thumb is that I will only denormalize a schema under limited circumstances:

  1. It has been demonstrated (and proven) that denormalization will resolve a performance issue.

  2. Storage of “static” (unchangeable) data. (e.g. Activity log, audit, and historical records)

  3. Storage of arbitrary document data as JSON or XML

The perceived complexity of writing a query is never considered as a basis for denormalization.