Redpill me on how can Google deliver millions of results within less than half a second
Redpill me on how can Google deliver millions of results within less than half a second
Nathaniel Walker
Other urls found in this thread:
Carson Walker
scroll down, it only delivers about 10 at a time. the rest is estimated
Landon Parker
>what are server farms
>what is virtualization
>what is scaling
>what is sharding
Zachary Barnes
>sharting
I do that everyday user, it's nothing special
Xavier Ortiz
Shading, you illiterate nigger
Brandon Myers
*brraaaaaaap*
Jaxson Foster
>Redpill me on how can Google deliver millions of results within less than half a second
by hiring the smartest people in the world and pay them well
Cooper Robinson
select shit from database where niggers LIMIT 10 OFFSET (page * 20);
select count(*) from database where niggers;
it's only delivering 10.
Liam Brooks
Thank you aha
Jaxon Rivera
>search for something
>we got 64m results!!
>go to page 2
>we got 500k results
>go to page 3
>actually this is the last page we only got 30 results
the current state of google
Easton Gray
It still has to scan a massive index to bring forth the 10 most relevant.
Carson Rodriguez
>brainlet detected
you haven't thought this through
Tyler Watson
Nice digits.
This is so fucking true, plus those 30 results are just like 3 fucking sites over and over again.
Noah Price
Text data does not take up as much space as images or videos, that said, there is still exabytes of text data to search through for a unique query.
The way it works, is when you send a query, say for a very specific string that might only exist on your website and you wrap it in quotations for example, the query is sent to (literally) hundreds of computers that all scan a different section of the entire database. The 10 most relevant results from each computer are sent to a table where it is again organized to give you the 10 most relevant results of those results.
So essentially, every time you make a query, you are harnessing hundreds of computers all at once. Google has multiple massive warehouses of computers to accomplish this. All funded through the billions they make in ad revenue.
Google also plays some tricks on you. For example, every keystroke you type in google is streamed before you even press search, and it searches through a cache of similar queries, so that result might already exist in a cache, this allows them to save computer resources.
Brayden Smith
They just sort by whoever paid them the most.
Justin Kelly
There are ways around that, but yes, they will certainly rig their own system to do that.
Adrian Collins
By using a highly optimized and efficient query language that traverses a knowledge graph.
Jonathan Young
Aaron Carter
>every keystroke you type in google is streamed before you even press search
Is Google keylogging me?
Ayden Lewis
yes
Thomas Barnes
they use mongodb it is web scale
Jack Green
fpbp
they don't need to prepare millions of results. They just need the top 10, and they'll prepare the next 50 or so as you keep clicking.
and they define top 10 to mean whatever the fuck they want it to mean. moving target.
Andrew Wright
using basic computer science and math techniques.
Justin Hernandez
it's probably not estimated, but it doesn't really matter either way.