HackerTube
HomeLongsLikedHistoryWatch laterCategoriesPlaylistsSettings

dotScale 2015 - John Graham-Cumming - I got 10 trillion problems, but logging ain't one

dotconferences1 HN recommender · recommended in 1 month, 2015-2018
YouTube
7K views · 11 years ago · talk · dated

John Graham-Cumming of Cloudflare talks at dotScale 2015 about handling roughly 10 trillion log lines a month. The talk touches on Cloudflare's hardware, its logging stack and its use of Lua in Nginx.

Mostly cited as a source for details of Cloudflare's architecture, such as LuaJIT-based request inspection and request volumes.

Cloud, DevOps & reliabilityNetworking & internet infrastructureDistributed systems#cloudflare#logging#lua#nginx#infrastructure

6 comments from Hacker News

@exikyut2018 · in “Ask HN: If I Close My Data Centers, What About the People/Jobs Lost?”

> Even putting aside that the computer tech part of the business isn't their core competency, custom hardware is hard. Hmmm. Good point. (Maybe I scared OP away, heh.) > Is there an example of any large company, other than Google, who has mastered this? Well, I certainly don't know. I've heard Cloudflare gets their systems made, for some definition of "from scratch", by Quanta. Presumably there's a bit of a decision-making process in there, and presumably Quanta handles everything related to the "why is the devkit sample broken again #51383" side of things and ships a more or less finished product (maybe even with finished enclosures). My reference point for this knowledge is ▶ this video (2015). Regarding SAS, maybe past a certain ordering quantity (or anything similar that would reveal a massive operation) the market would expect certain behavior like requesting value-adds, ordering with only one supplier, or other things that would generate lock-in. I of course have absolutely no idea [if this sort of thinking is reasonable]. Your points about commodity at scale sound quite reasonable. (You clearly have more experience than I do, which is kind of zero, heh.) As for suspicion about summaries and conclusions, well, (tangential example) the radial graph on link (the one from Forrester) doesn't show GitHub :D Yeah, building own hardware is sadly still within the realm of over-specialization at this point. I guess this is because 4+-layer PCB printing is still eye-watering so the

CitesView on HN
@Artemis22017 · in “Quantifying the Impact of “Cloudbleed””

For reference, two years ago Cloudflare was serving more than three million requests per second[0]. I can only imagine this figure has gone up a lot. 0: ▶ this video

CitesView on HN
@i336_2016 · in “Hello Lua”

You make incredibly fair points and I don't have the data to answer what you've raised. This could be lack of research or the fact that CF hasn't released it. I expect a 20/80 non-researched/non-available mix. My understanding is that the WAF detects trivial stuff like SQL injection and XSS, and the video also mentions it detected ShellShock and so forth (yeah, this was a while ago). "Thousands of regular expressions" is mentioned verbally in the video around ~11:25. That's where I got that from. There are some bits mentioned about the hardware around 19:30: - E5-2630v3 (16 cores, 32 w/ HT which is turned off), 128GB RAM, 10Gbps - Custom built by Quanta: 4 nodes per 2U (each node is like half-width and 1U high), redundant 2xPSU in the middle - 40 machines run Kafka storing 400TB of log data (ingest @ 15Gbps), 50TB disk per node - 5 machines run CitusDB (sharded PostgreSQL) to power realtime graph generation, 12TB SSD per node - 100+ machines for realtime analytics running Go (and presumably LuaJIT) - LuaJIT seems to be the realtime "hmm, interesting" flagging system, while Go is used for in-depth intelligent analysis that's (I assume) slightly behind realtime. Also, the video is actually over here - ▶ this video --- I'm curious to understand what tail latency is, and very interested to learn more about language latency and performance in general. What subjects would you recommend I learn about / read? (I would describe myself as interested from a hobbyist-intermediate standpont, b

CitesView on HN
@i336_2016 · in “Ask HN: Whats your favorite HN thread from 2015?”

I discovered a moderately old video from CloudFlare hiding in there, and had a rather big TIL moment. It's one thing to read "LuaJIT is one of the fastest dynamic languages" etc blah blah on link ... ...it's quite another to learn that EVERY SINGLE CLOUDFLARE REQUEST - all 4 million/sec of them - goes through a LuaJIT-based attack detection system. Oh. Video: ▶ this video (from link ) Found in the linked thread here: link

RecommendsView on HN
@mtourne2015 · in “OpenResty – A fast web app server by extending Nginx”

Rewriting the hodge podge of C, Perl and other custom bits of Nginx almost entirely in Lua and hiring its maintainer Yichun in the process where important steps for scaling up the CloudFlare's CDN and make sure each node is running as hot as possible. More on this from dotScale 2015: ▶ this video

CitesView on HN
@mdaniel2015 · in “How CloudFlare extracts a signal from 10 trillion log lines each month [video]”

What is the deal with the recent "thedotpost.com" envelope for perfectly valid YouTube URLs? I want to flag this post just for that. ▶ this video

CriticalView on HN
Rick Reed - WhatsApp: Half a billion unsuspecting FreeBSD users54:14

Rick Reed - WhatsApp: Half a billion unsuspecting FreeBSD users

iX
31K views • 11 years ago
Seattle Conference on Scalability: YouTube Scalability52:41

Seattle Conference on Scalability: YouTube Scalability

Google TechTalks
30K views • 19 years ago
OpenFlow @ Google - Urs Hoelzle, Google47:18

OpenFlow @ Google - Urs Hoelzle, Google

Open Networking Summit
47K views • 14 years ago
Inside a Google data center5:28

Inside a Google data center

Google Workspace
22.6M views • 11 years ago
NYC Tech Talk Series: How Google Backs Up the Internet1:14:19

NYC Tech Talk Series: How Google Backs Up the Internet

Google TechTalks
41K views • 12 years ago
AWS re:Invent 2014 | (SPOT302) Under the Covers of AWS: Its Core Distributed Systems46:15

AWS re:Invent 2014 | (SPOT302) Under the Covers of AWS: Its Core Distributed Systems

Amazon Web Services
10K views • 11 years ago
High Volume, Bandwidth Heavy Infrastructure On The Cheap - Kyle Drake46:23

High Volume, Bandwidth Heavy Infrastructure On The Cheap - Kyle Drake

PDX Node
1K views • 9 years ago
AWS re:Invent 2014: AWS Innovation at Scale with James Hamilton (SPOT301)47:51

AWS re:Invent 2014: AWS Innovation at Scale with James Hamilton (SPOT301)

Amazon Web Services
70K views • 11 years ago
Scaling the Linear Sync Engine1:06:08

Scaling the Linear Sync Engine

Linear
37K views • 3 years ago
How We've Scaled Dropbox1:08:17

How We've Scaled Dropbox

Stanford
372K views • 14 years ago
Scalability at YouTube38:43

Scalability at YouTube

Next Day Video
51K views • 14 years ago
Putting Google's Network to work for You7:06

Putting Google's Network to work for You

Google for Developers
23K views • 12 years ago
AWS re:Invent 2016: Tuesday Night Live with James Hamilton1:33:27

AWS re:Invent 2016: Tuesday Night Live with James Hamilton

Amazon Web Services
133K views • 9 years ago
Velocity 2012:  Richard Cook, "How Complex Systems Fail"28:21

Velocity 2012: Richard Cook, "How Complex Systems Fail"

O'Reilly
60K views • 14 years ago
Building Software Systems At Google and Lessons Learned1:22:45

Building Software Systems At Google and Lessons Learned

Stanford
257K views • 15 years ago
"Stop Rate Limiting! Capacity Management Done Right" by Jon Moore42:20

"Stop Rate Limiting! Capacity Management Done Right" by Jon Moore

Strange Loop Conference
41K views • 9 years ago
AWS re:Invent 2018: How AWS Minimizes the Blast Radius of Failures (ARC338)55:44

AWS re:Invent 2018: How AWS Minimizes the Blast Radius of Failures (ARC338)

Amazon Web Services
37K views • 7 years ago
dotScale 2014 - Robert Kennedy - Life in the Trenches of healthcare.gov27:21

dotScale 2014 - Robert Kennedy - Life in the Trenches of healthcare.gov

dotconferences
13K views • 12 years ago
"Systems that run forever self-heal and scale" by Joe Armstrong (2013)1:10:23

"Systems that run forever self-heal and scale" by Joe Armstrong (2013)

Strange Loop Conference
85K views • 5 years ago
Google I/O 2012 - Go Concurrency Patterns51:27

Google I/O 2012 - Go Concurrency Patterns

Google for Developers
894K views • 14 years ago