Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How Search Works

google.com

Post reply on HN
Print view View original

91–100 of 108 posts

Re: How Search Works

#91

Post by jfim »

alok-g wrote: post #24
manojlds wrote: post #18

> By the way, in the 47 seconds you've been on this page, approximately 1,813,260 searches were performed. Aren't these just some random numbers that they pull out of the air?

That would be about 38K searches per second. Does this include Google instant searches? Google search results show a time value for each search. E.g.: About 2,210,000,000 results (0.12 seconds). Is this time machine time per search? This number is often around 30 ms, give or take a factor of two. If so, each machine can handle about 30 searches per second. If so, 38K searches per second need about 1000 machines. Soun…

It's probably the wall time for the various backend services to respond to the query. If you think about it, a Google search result is actually many things; it has results from various sources, such as the web, images, videos, news, social signals from G+, etc. All of those are different services that are aggregated to build your result page.

Since all of those queries are fired at the same time, the only metric that matters at the end is the wall time, not the CPU time used during the query.

I also seriously doubt that the servers that handle the Google front page can only do one query at a time; at the very least, they're multithreaded, but probably concurrent. It probably works as below:

1. Parse query 2. Send query to backend servers 3. Wait until all backends replied or at most 250ms (or some other timeout) 4. Assemble the result page and ship it back to the client

While the server is idling for the backends to reply, it probably processes other queries; it wouldn't make sense to waste that much CPU power.

Finally, your example says 0.12s (a random query on my end gave a response time of 0.69s), which is 120ms (or 690ms for mine), which is more than twice 30ms.

Re: How Search Works

#92

Post by lucb1e »

gnu8 wrote: post #35
aviswanathan wrote: post #10

Scrolling is really becoming the new thing in UX design. It's an interesting contrast to the 'movie-like' flash animations of a few years ago that required no interaction on behalf of the user.

At some point users started closing the tab instantly as soon as it becomes clear something non-interactive like a flash movie is the central element of the page. The scrolling page is the optimal way of letting the user read the content at the desired pace.

A youtube clip would have done just fine. You can pause that if you need a slower pace. Scrolling is very slow here (and I'm using Google's own Chrome on mainstream hardware), and it's never clear where I should stop scrolling to view the page. I don't get the fuzz about scrolling websites.

Re: How Search Works

#94

Post by Inufu »

Julianhearn wrote: post #90
franze wrote: post #26

thx matt and the google search team for doing this. it's nothing new for technically inclined people, but every little bit helps. helps for what? teaching people to worry about the right aspects of search and the impact on their business, instead of worrying about bullshitphrases that were planted in their head by a SEO agency key account or a blogpost from 2008. so well yes, thx for doing this. i will send it to my…

Your 80% is based on what exactly? A tiny sample size. Please if you don't have solid data don't quote percentages it just encourages people to spread the number like it's a fact, which it isn't. If you read the right sources a majority of seo advice is correct. Www.seomoz.org http://static.googleusercontent.com/external_content/untrust... Www.inbound.org (homepage stuff that has been voted up.)

>> If you read the right sources a majority of seo advice is correct.

That's a contradiction. If you have to read the right sources, then by definition the majority of advice is not correct.

Re: How Search Works

#95

Post by SquareWheel »

JDDunn9 wrote: post #63
moultano wrote: post #29

Earlier quoted context omitted.

This is completely false. The effect on revenue is not used to make launch decisions for ranking changes.

So when Google's Panda update killed tons of user-generated-content sites like Mahalo, eHow, HubPages, etc., and greatly improved YouTube's (which is 99% garbage) rankings, that was pure co-incidence? What about when Google rolled out universal search only after buying YouTube?

    "What about when Google rolled out universal search only after buying YouTube?"
Because the technology wasn't built then? Google had a video platform before they owned Youtube, you know.

Re: How Search Works

#96

Post by bambax »

dylangs1030 wrote: post #88

I don't know what to take from this. That search is very complex (I knew that, but not with this technical detail). Or ...that Google is trying very hard to maintain user interest with gimmicky shows of why it's cool and cutting edge and necessary. Not that Google isn't those things...this just seems like an unnecessary expenditure of time. We know it's complex Google. Improve some other features and stop shutting ot…

I too find this completely void of any useful information.

There are many things that are still broken in search; I talk about one specific experience here:

http://urgeous.com/p87t3aaa40g-for-some-queries-all-first-10...

("For some queries, all first 10 results on Google are spam").

Re: How Search Works

#97

Post by Julianhearn »

Inufu wrote: post #94
Julianhearn wrote: post #90

Earlier quoted context omitted.

Your 80% is based on what exactly? A tiny sample size. Please if you don't have solid data don't quote percentages it just encourages people to spread the number like it's a fact, which it isn't. If you read the right sources a majority of seo advice is correct. Www.seomoz.org http://static.googleusercontent.com/external_content/untrust... Www.inbound.org (homepage stuff that has been voted up.)

>> If you read the right sources a majority of seo advice is correct. That's a contradiction. If you have to read the right sources, then by definition the majority of advice is not correct.

Why does it mean the majority of advice is not correct? That is a myth.

Re: How Search Works

#98

Post by ScottWhigham »

crynix wrote: post #3
area51mafia wrote: post #2

It's nice overall, but the timing for making items appear is a little slow. I was past most headers by the time they appeared, and I don't think I scroll too incredibly fast.

It might be your machine/browser/os. The same thing happened on my laptop (OSX/Chromium), but it worked fine on my desktop (Arch/Chromium).

If you read the top voted comment for this post (as I write this), it describes the same experience though.

Re: How Search Works

#99

Post by eitland »

moultano wrote: post #87
_mvuc wrote: post #60

I keep checking every so often, but searching for "this phrase" or +absolute +requirement is still broken. Even "Verbatim", isn't. If they can't even get simple search right, who would trust them with anything more?

Do you have some example queries to debug?

I have: - http://techinorg.blogspot.com/2013/03/what-is-going-on-with-...

edit: Here are just the queries:

"sublime text 2" "focus group"

cisco "anyclient" - this one gets silently rewritten to cisco anyconnect

shopify "deduplicate" - with verbatim activated -

Re: How Search Works

#100

Post by eitland »

_mvuc wrote: post #60

I keep checking every so often, but searching for "this phrase" or +absolute +requirement is still broken. Even "Verbatim", isn't. If they can't even get simple search right, who would trust them with anything more?

Same here. Tried reporting it a number of times in different ways but nothing ever happens. Have posted a few examples as response to moultanos answer further down in this thread.
Post reply on HN