Learn Database Denormalization

CORRECTIONS: Expand this comment to see corrections. * CORRECTIONS * At 12:00, a couple of the columns on the Subtask table are oddly named. "Task_Assignment_Datetime" and "Task_Type_Code" should really have been called "Subtask_Assignment_Datetime" and "Subtask_Type_Code", respectively. In addition, one of the assignment datetimes (November 12th at 204pm) is unrealistic, as we'd expect it to be later than the creation datetime of the parent Task.

The return of the king (2022)

As a database administrator, I approve of your explanations. I work with tons of data, and I wish every database was normalized and had solid loads for denormalized reporting databases. But in the real world it is rare. When we get feeds from outside and then need to report on them it always makes the process ridden with issues. Unfortunately having 10 different data sources outside of your control and then trying to join them for reporting is why database administration is still a solid job.

One project, we deliberately did a 'denormalizing' thing. We had data from instruments that took readings every 5 minutes and recorded this raw data to a table. Then, we had to get an accumulated calculation that basically summed these readings for an entire month, quarter, and year (well, it was more complex than that, but that's the general idea). These 'sums' had to be accessed often for various reasons. So we made a table of 'cached sums' where we saved the results of the 'expensive, time-consuming' calculation. If there was no table entry for a particular instrument, we ran the 'slow' calculation that queried every instrument reading for the past year, saved the result in the 'cached sums' table and returned the result. Next time we needed that particular sum, it would be quickly returned from the 'cached sums' table. So the 'cached sums' data was, technically speaking, redundant data (the original readings being the 'source data'). We did consider that editing the raw readings causes the 'cached sums' to be invalidated, but we compromised.

Love the content! Helped me get a job. Great explanations and practical examples! Please upload more - I think this channel could be huge over the long run!

As a junior developer who didn’t have enough knowledge about databases, I stumbled upon this video because I was curious about ‘What is Denormalization?’ Your video introduced me to many new concepts that I now need to contemplate and learn. Thank you very much.

That was a very nice example with the 'UnitPrice'. Ultimately it's about knowing your domain and distinguishing , what Uncle Bob calls, 'critical business rules' from something rather temporary and not so 'critical'.

This is one of a simplest explanation for a complicated topic. Good work, Decomplexify! Living up to the name.

Always happy to see new content from you!

I misunderstood the topic, I thought it‘s about „reverse-engeneering“ a given (older) database. But it was great to watch. Vocabulary is everything! Thank you so much for your efforts!!! ❤️❤️❤️

Your channel is solid and informative! Please keep making content!

Superb examples and explanation in practical terms!

Please keep making videos! They are so good!

Amazing explanation covering all the denormalization facets. Congrats

You have THE BEST database videos. Like... Ever.

Thank you SO MUCH for such neat, organized and really SIMPLE to understand explanations! Not just this but I learned a lot from most of your videos. Keep up the great work and thanks a lot (Y)

Thanks for making this. I think I requested this on your normalization video - appreciate it!

I should be paying my tuition to you instead, but all I can thank you with in this instance is a thank you and a like button. Thank you.

Thank you for your amazing videos !

New decomplexify video!! Great stuff