That's a lot. With Percona clusters I started having issues requiring fine-tuning around a third of that at quite short peak loads, maybe ten minutes sustained high load topping out at 6-10k writes/s. Something like 24 cores, 192 GB RAM on the main node.
Not sure how GC works in Golang but if you see 20k writes/s sustained that's what I'd be nervous about. If every write is 4 kB I think it would be something like a quarter of a TB per hour, probably a full TB at edge due to HTTP overhead, so, yeah, a lot to handle on a single node.
Maybe there are performance tricks I don't know about that makes 20k sustained a breeze, I just know that I had to spend time tuning RAM usage and whatnot for peaks quite a bit earlier and already at that load planned for sharding the traffic.
“Teasers are usually rich kids with nothing to do. They cruise around looking for planets that haven’t made interstellar contact yet and buzz them.”
“Buzz them?” Arthur began to feel that Ford was enjoying making life difficult for him.
“Yeah,” said Ford, “they buzz them. They find some isolated spot with very few people around, then land right by some poor unsuspecting soul whom no one’s ever going to believe and then strut up and down in front of him wearing silly antennas on their head and making beep beep noises.”
My three year old and I actually had a lot of fun with the Calculator app on my phone just a couple hours ago. He thought that me reading off large numbers was hilarious.
So, yeah, Calculator is a pretty important feature.
Full decode of the P section is even more amusing:
GEEK PERL CODE [P+++++(--)$]
My tendencies on this issue range from: "I am Larry Wall, Tom Christiansen, or Randal Schwartz.", to: "Perl users are sick, twisted programmers who are just showing off." Getting paid for it!
Needed more practice in construction estimating, which is yet another discipline that you wouldn’t immediately think of when someone says “build a house.”
A friend of a friend runs a pipeline construction company (think the people that would actually build Keystone XL or the likes) and he has a team of estimators that run all the numbers on a given project and come up with a budget. He then double checks their numbers by chartering a helicopter and flying the route with an an old construction foreman, who basically eyeballs the thing and and does some mental math. Apparently it was uncanny how often the two estimates were within 10% of each other, but that the old guy eyeballing the thing had saved him on a number of occasions by recognizing issues/costs that the estimators had missed...
After spending his entire career in construction, one of the last jobs my dad had before retiring was in preparing bids for proposals. The company wanted people with lots of experience in the bidding process. Not just on-budget, on-schedule, but people that had been through over-budget, behind-schedule so that they knew what to be aware of as potential pit falls.
When I was in college one of the big promises of BIM was "we'll have information rich models that will make estimating easier and more accurate!"
I'm not directly involved in that process in-industry (ended up in light fixture manufacturing), but my impression is that BIM models are usually good enough to kick out a set of drawings, but you wouldn't want to count on any information that you pull out of it electronically being accurate.
So you still have "contractor to verify quantity" notes on everything, and presumably someone going through the drawing figuring out how many "Type F 2x4' troffers" and how many faucets and how many of every other little detail that will take money or time.
It's a pretty intensive process, but given the amounts of money at stake if you screw it up, hugely important to get right.
Software development is more like - couple hundred percent over spec in time and cost, code base is a mess, almost nobody uses it and finally, after a couple of years, management changes and the new mgmt calls a spade a spade and terminates the project.
At least that's often the case in huge organizations trying to do crazily ambitious projects (esp. given their general incompetence). I tend to consult on such shitstorms because it pays well. The scars are for life though...
