How I Learned to Stop Worrying and Love Using a Lot of Disk Space to Scale
You think you know database design? Think again!
|
How I Learned to Stop Worrying and Love Using a Lot of Disk Space to ScaleYou think you know database design? Think again!
1 Comment » |
I would love this, but currently one of the constraint on the DB is disk space - I know, hard to believe, but that’s true. Disk space is something we cannot waste.
Incidentally denormalization practices date back to the 70s, at least, so nothing new there
“Disk and CPU are cheap so stop worrying about them and scale.”
Except, disk is not cheap. Not neccessarily. If you are in some RAID/BLADE environment, then you can end up with costs in the N*1000 USD range.
For some companies that’s a lot - whether they are right or not, I will not go into, I just note that “disk is cheap” is not entirely true - and if you happen to work at such a company you can’t go with the “disk is cheap” mentality.
I learned to “love” the, “hey, don’t think, just buy more stuff that will solve it” mentality. I think we should adopt it all, and the Computer Science as is should go to the level of woodcutters and bricklayers in needs of education, and salary of course.
The article is a good one and a valid one in many points, and for some problems. But I am afraid it will end up with a short summary of “no design, buy hardware” in most of the heads and that would be bad.
If you think I am wrong, well, I hope I am, but just look at the “Agile” methods as used. Practically rush to code at most of the companies I saw.