So, recently I was wondering how much can a horse carry. The recommendation is to not to load a horse by more than 20% of their body weight. So, on average a horse is around 1200 pounds and the limit is usually around 240 pounds.
Now, we’re talking about beasts of burden and therefore referring to draft horses.
This is the limitation of this model.
If we use the methodology of using a cart, then each horse can actually pull around 8000 pounds increasing your productivity by 97%.
It’s still a horse based model and there’s still just one horse. But, the methodology was more important. So, when people tell you, just use the model and not the methodology with the Data Vault, then you should raise your eyebrows (At least one of them).
But, lets not stop here …
Let’s start optimizing the other portions and add parallelism like Data Vault 2.0 because 1.0 can’t parallelize loads across all object types, but 2.0 can.
Let’s go back to the horses.
Do you know how much 2 horses can pull?
If you’re thinking, the same or double, then you’d be stumped when you hear this. Two of them can pull up to 24000 pounds. That is 3 times their single capacity.
And, that’s just by adding a new methodology.
But, that’s not all …
Lets add a process architecture where the horses are trained together and have worked together before. Then they can pull a combined weight of a staggering 32000 pounds which is 4 times of what they can each do.
Data Vault 2.0 is about people, process and technology and optimization of everything. The model is the simplest of this all, and only enables optimization to be leveraged by the processes. For example, using hashes instead of numeric surrogates enables parallel loading of differing objects like hubs and links and even their satellites. But, the model doesn’t really tell you to do so. You can still load them sequentially. You should follow the established methodology to get the most out of it.
So, like the horses, just switching the DV 1.0 to DV 2.0 model may get you a very small level of gain, but you’re still missing out on a majority of the benefits when using only the model. It’s the other parts when used with the model give you the biggest bank for your buck … so to speak.
Dan wrote about this very thing here and it’s worth a read:
=> https://www.linkedin.com/pulse/data-vault-20-beyond-model-dan-linstedt/
According to Dan’s estimates, focusing purely on the model restricts you to only 10% of the benefits you can get from DV 2.0. Don’t let anyone fool you on this, because he invented both DV 1.0 and 2.0 and he himself is recommending 2.0 now, which would be reason enough to look at it. He could have continued teaching and certifying people in DV 1.0, but he chose not to. Something to keep in mind.
The horse example also works great to illustrate a trained agile team v/s just throwing in agile as a buzzword which I have seen before. DV 2.0 incorporates something called Disciplined Agile Deliveries which in my opinion is an optimal and pragmatic approach to agile. It’s why even Scott Ambler endorses DV 2.0.
Reference -> http://timmaurer.com/2012/01/16/horse-sense/