The C10M problem

#1

Post by z_ » Mon, Feb 17, 2014, 3:32 AM UTC

The C10M problem
c10m.robertgraham.com

Re: The C10M problem

#2

Post by jared314 » Mon, Feb 17, 2014, 4:18 AM UTC

Previous Discussion: https://news.ycombinator.com/item?id=5699552 (9 months ago)

High Scalability Post: http://highscalability.com/blog/2013/5/13/the-secret-to-10-m...

Original Shmoocon Presentation: http://www.youtube.com/watch?v=73XNtI0w7jA

Re: The C10M problem

#3

Post by BadassFractal » Mon, Feb 17, 2014, 4:30 AM UTC

This article on High Scalability also covers part of the problem: http://highscalability.com/blog/2014/2/5/littles-law-scalabi...

Re: The C10M problem

#4

Post by ksec » Mon, Feb 17, 2014, 4:30 AM UTC

I think OSv or something similar would be part of that solution. Single User / Purpose OS designed to do one / few things and those only.

I could only hope OSv development would move faster.

Re: The C10M problem

#5

Post by ehsanu1 » Mon, Feb 17, 2014, 4:33 AM UTC

An implementation of the idea: http://www.openmirage.org/

A good talk about it by one of the developers/researchers: http://vimeo.com/16189862

Re: The C10M problem

#6

Post by nwmcsween » Mon, Feb 17, 2014, 4:34 AM UTC

So an exokernel?

Re: The C10M problem

#7

Post by voltagex_ » Mon, Feb 17, 2014, 4:37 AM UTC

>Content Blocked (content_filter_denied)

>Content Category: "Piracy/Copyright Concerns"

I'm starting to use these blocks at my workplace as a measure of site quality (this will be a high quality article). Can someone dump the text for me?

Re: The C10M problem

#8

Post by toomuchtodo » Mon, Feb 17, 2014, 4:39 AM UTC

voltagex_ wrote: post #7

>Content Blocked (content_filter_denied) >Content Category: "Piracy/Copyright Concerns" I'm starting to use these blocks at my workplace as a measure of site quality (this will be a high quality article). Can someone dump the text for me?

Does it work through IA?

http://web.archive.org/web/20140217043940/http://c10m.robert...

Re: The C10M problem

#9

Post by sb057 » Mon, Feb 17, 2014, 4:40 AM UTC

voltagex_ wrote: post #7

>Content Blocked (content_filter_denied) >Content Category: "Piracy/Copyright Concerns" I'm starting to use these blocks at my workplace as a measure of site quality (this will be a high quality article). Can someone dump the text for me?

That page is essentially a glorified intro to his series of blog entries, and they are on another domain, so perhaps they are not blocked:

http://blog.erratasec.com/search/label/C10M

Re: The C10M problem

#10

Post by rdtsc » Mon, Feb 17, 2014, 4:48 AM UTC

Here is how C2Mhttp://www.erlang-factory.com/upload/presentations/558/efsf2...

It shows good practical tricks and pitfalls. It was 3 years ago so I can only assume it got better, but who knows.

Here is the thing though, do you need to solve C*M problem on a single machine? Sometimes you do but sometimes you don't. But if you don't and you distribute your system you have to fight against sequential points in your system. So you put a load balancer and spread your requests across 100 servers each 100K connections. Feels like a win, except if all those connections have to live at the same time and then access a common ACID DB back-end. So now you have to think about your storage backend, can that scale? If your existing db can't handle, now you have to think about your data model. And then if you redesign your data model, now you might have to redesign your application's behavior and so on.