Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Building an automated DDoS mitigation pipeline
Search
majek04
January 23, 2016
Programming
5
1.3k
Building an automated DDoS mitigation pipeline
majek04
January 23, 2016
Tweet
Share
More Decks by majek04
See All by majek04
Keeloq_AD_2025__1_.pdf
majek04
0
18
BPF programmable socket lookup
majek04
0
650
Linux at Cloudflare
majek04
3
8.5k
DDoS Landscape
majek04
0
420
Inside Cloudbleed
majek04
3
2.9k
Golang sucks
majek04
21
52k
Gatelogic - Somewhat functional reactive framework in Python
majek04
1
5.1k
How Cloudflare deals with largest DDoS attacks?
majek04
2
3.5k
Why we chose Service Worker API
majek04
0
2.7k
Other Decks in Programming
See All in Programming
【卒業研究】会話ログ分析によるユーザーごとの関心に応じた話題提案手法
momok47
0
130
C-Shared Buildで突破するAI Agent バックテストの壁
po3rin
0
420
Grafana:建立系統全知視角的捷徑
blueswen
0
230
Context is King? 〜Verifiability時代とコンテキスト設計 / Beyond "Context is King"
rkaga
10
1.4k
フルサイクルエンジニアリングをAI Agentで全自動化したい 〜構想と現在地〜
kamina_zzz
0
310
LLMで複雑な検索条件アセットから脱却する!! 生成的検索インタフェースの設計論
po3rin
4
990
Navigation 3: 적응형 UI를 위한 앱 탐색
fornewid
1
470
PostgreSQLで手軽にDuckDBを使う!DuckDB&pg_duckdb入門/osc25hi-duckdb
takahashiikki
0
190
Implementation Patterns
denyspoltorak
0
130
「コードは上から下へ読むのが一番」と思った時に、思い出してほしい話
panda728
PRO
39
26k
AIコーディングエージェント(NotebookLM)
kondai24
0
240
Cell-Based Architecture
larchanjo
0
150
Featured
See All Featured
[RailsConf 2023 Opening Keynote] The Magic of Rails
eileencodes
31
9.8k
技術選定の審美眼(2025年版) / Understanding the Spiral of Technologies 2025 edition
twada
PRO
115
94k
Jess Joyce - The Pitfalls of Following Frameworks
techseoconnect
PRO
1
31
KATA
mclloyd
PRO
33
15k
Redefining SEO in the New Era of Traffic Generation
szymonslowik
1
170
The AI Revolution Will Not Be Monopolized: How open-source beats economies of scale, even for LLMs
inesmontani
PRO
3
2.8k
Rebuilding a faster, lazier Slack
samanthasiow
85
9.3k
Are puppies a ranking factor?
jonoalderson
0
2.5k
Code Reviewing Like a Champion
maltzj
527
40k
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
508
140k
Mind Mapping
helmedeiros
PRO
0
40
Believing is Seeing
oripsolob
0
16
Transcript
Building an automated DDoS Mitigation Pipeline Marek Majkowski
2 "Help Build a Better Internet"
Content neutral 3
DDoS is a threat 4
5 Malicious Attacker Internet Provider Origin Server CloudFlare Server trust
& safety team w orking w ith operators public outreach Big effort im proving our infrastructure
6 Automated DDoS Mitigations Malicious Attacker Internet Provider Origin Server
CloudFlare Server autom ating m itigations
7 attack volume CloudFlare network capacity >
BGP Nullroute and move on 8 ! route 1.2.3.4/32 {!
discard;! community [ 13335:666 13335:668 13335:36006 ];! }!
attack volume CloudFlare network capacity < 9
10 BGP Nullrouting Router firewall Server firewall Application Less damage
Reducing damage
11 BGP Nullrouting IP Router firewall IP, port, packet length
Server firewall all above + stateless DPI parameters Application all above + application logic More precision Reducing damage
12 Operator Precision Speed
13
14 Automation Precision Speed
15 Gatebot Precision Speed Automatic attack handling
Attack Detection Automatic attack handling 16 Mitigation Reactive Automation
The attack 17
High volume packet floods 18 Packets per second
DNS packet flood 19 ! $ tcpdump -ni eth2 inbound
and port 53 -c 100! ! IP 202.194.181.95.15443 > 1.2.3.4:53: 63476% [1au] A? example.com. (50)! IP 221.12.236.115.6570 > 1.2.3.4:53: 11406% [1au] A? example.com. (50)! IP 203.94.134.43.18473 > 1.2.3.4:53: 8559% [1au] A? example.com. (50)! IP 203.196.66.75.32573 > 1.2.3.4:53: 47971% [1au] A? example.com. (50)! IP 124.240.198.136.2333 > 1.2.3.4:53: 61152% [1au] A? example.com. (50)! IP 218.247.70.185.11679 > 1.2.3.4:53: 16360% [1au] A? example.com. (50)! IP 202.109.218.98.27549 > 1.2.3.4:53: 17829% [1au] A? example.com. (50)! IP 203.148.240.82.21825 > 1.2.3.4:53: 22590% [1au] A? example.com. (50)! IP 211.167.108.67.25782 > 1.2.3.4:53: 17663% [1au] A? example.com. (50)! IP 203.209.60.18.20221 > 1.2.3.4:53: 38257% [1au] A? example.com. (50)! IP 203.81.181.168.12749 > 1.2.3.4:53: 53492% [1au] A? example.com. (50)!
1 in 10k packets is "real" 20
Finding attack parameters 21 ! IP 202.194.181.95.15443 > 1.2.3.4:53: 63476%
[1au] A? example.com. (50)! IP 221.12.236.115.6570 > 1.2.3.4:53: 11406% [1au] A? example.com. (50)! IP 203.94.134.43.18473 > 1.2.3.4:53: 8559% [1au] A? example.com. (50)! IP 203.196.66.75.32573 > 1.2.3.4:53: 47971% [1au] A? example.com. (50)! IP 124.240.198.136.2336 > 1.2.3.4:53: 61152% [1au] A? example.com. (50)! IP 218.247.70.185.11679 > 1.2.3.4:53: 16360% [1au] A? example.com. (50)! IP 202.109.218.98.27549 > 1.2.3.4:53: 17829% [1au] A? example.com. (50)! IP 203.148.240.82.21825 > 1.2.3.4:53: 22590% [1au] A? example.com. (50)! IP 211.167.108.67.25782 > 1.2.3.4:53: 17663% [1au] A? example.com. (50)! IP 203.209.60.18.20221 > 1.2.3.4:53: 38257% [1au] A? example.com. (50)! IP 203.81.181.168.12749 > 1.2.3.4:53: 53492% [1au] A? example.com. (50)!
Mitigation 22 Mitigation Operator
Where to DROP? 23 Application iptables Router
Traffic matching with BPF 24 ! iptables -A INPUT \!
--dst 1.2.3.4 \! -p udp --dport 53 \! -m bpf --bytecode "14,0 0 0 20,177 0 0 0,12 0 0 0,7 0 0 0,64 0 0 0,21 0 7 124090465,64 0 0 4,21 0 5 1836084325,64 0 0 8,21 0 3 56848237,80 0 0 12,21 0 1 0,6 0 0 1,6 0 0 0" \! -j DROP!
25 ! ldx 4*([14]&0xf)! ld #34! add x! tax! lb_0:!
ldb [x + 0]! add x! add #1! tax! ld [x + 0]! jneq #0x07657861, lb_1! ld [x + 4]! jneq #0x6d706c65, lb_1! ld [x + 8]! jneq #0x03636f6d, lb_1! ldb [x + 12]! jneq #0x00, lb_1! ret #1! lb_1:! ret #0! BPF bytecode
26
Deployment 27 iptables Mitigation Database
Mitigation database 28 ! $ gatekeeper dnsbpf list! --ip=1.2.3.4 *.example.com!
--ip=4.3.2.1 www.test.de *.www.test.de! --ip=4.3.4.4 *.cloudflare.com --except=www.** --except=ns1.**! --ip=2.3.1.4 www.onedomain.com,wwww.seconddomain.com! --ip=1.2.3.0/24 test.com! ! $ gatekeeper dnsbpf add -- --ip=4.3.2.1 *.newattack.com!
Detection 29 Attack Detection
Sflow 30 Sflow Central Aggregation
What is an "attack"? 31
"Attack" is large 32 Large attacks Small attacks Packets per
second
33 Attacks Mitigation "Attack" can be mitigated Attack Detection Mitigation
Database Attack Description = Mitigation 33 iptables Sflow
34 ! Mpps Descr! 3.878 --ip=141.245.59.191/32! 2.878 --ip=141.245.59.192/32! 1.878 --ip=141.245.59.193/32!
1.878 --ip=141.245.59.194/32! 1.878 --ip=141.245.59.195/32! 1.878 --ip=141.245.59.196/32! 1.878 --ip=141.245.59.197/32! 1.878 --ip=141.245.59.198/32! 1.878 --ip=141.245.59.199/32! ...! ! Mpps Descr! 35.878 --ip=141.245.59.0/24! vs "Attacks" shall be aggregated
35 An attack-finding algorithm
Top N / Heavy hitters • Fixed memory size; Algorithm:
Space Saving • https://github.com/cloudflare/golibs 36 pps IP 12.2M 1.2.3.4 2.4M 42.1.2.4 0.01M 2.4.3.1 0.01M 192.168.1.1
Multiple dimensions 37 pps IP:port 12.2M 1.2.3.4:53 2.4M 42.1.2.4:80 0.01M
2.4.3.1:80 0.01M 192.168.1.1:443 pps IP 12.2M 1.2.3.4 2.4M 42.1.2.4 0.01M 2.4.3.1 0.01M 192.168.1.1 pps subnet 12.2M 1.2.3.0/24 2.4M 42.1.2.0/24 0.01M 2.4.3.0/24 0.01M 192.168.1.0/24
Multiple dimensions 38 pps IP:port 12.2M 1.2.3.4:53 2.4M 42.1.2.4:80 0.01M
2.4.3.1:80 0.01M 192.168.1.1:443 pps IP 12.2M 1.2.3.4 2.4M 42.1.2.4 0.01M 2.4.3.1 0.01M 192.168.1.1 pps subnet 12.2M 1.2.3.0/24 2.4M 42.1.2.0/24 0.01M 2.4.3.0/24 0.01M 192.168.1.0/24 incoming sample: 42.1.2.4:80
Multiple dimensions 39 pps IP:port 12.2M 1.2.3.4:53 2.4M 42.1.2.4:80 0.01M
2.4.3.1:80 0.01M 192.168.1.1:443 pps IP 12.2M 1.2.3.4 2.4M 42.1.2.4 0.01M 2.4.3.1 0.01M 192.168.1.1 pps subnet 12.2M 1.2.3.0/24 2.4M 42.1.2.0/24 0.01M 2.4.3.0/24 0.01M 192.168.1.0/24 reporting threshold: 1M
Attack report 40 ! Mpps Descr! 12.2 --ip=1.2.3.4 --port=53! 2.4
--ip=42.1.2.4 --port=80! 12.2 --ip=1.2.3.4! 2.4 --ip=42.1.2.4! 12.2 --ip=1.2.3.0/24! 2.4 --ip=42.1.2.0/24!
Multiple dimensions 41 pps IP:port 12.2M 1.2.3.4:53 2.4M 42.1.2.4:80 0.01M
2.4.3.1:80 0.01M 192.168.1.1:443 pps IP 0.1M 1.2.3.4 0M 42.1.2.4 0.01M 2.4.3.1 0.01M 192.168.1.1 pps subnet 0.1M 1.2.3.0/24 0M 42.1.2.0/24 0.01M 2.4.3.0/24 0.01M 192.168.1.0/24 incoming sample: 42.1.2.4:80
Attack report 42 ! Mpps Descr! 12.2 --ip=1.2.3.4 --port=53! 2.4
--ip=42.1.2.4 --port=80!
Scales well 43
Reactive automation 44 Reactive Automation
Connecting the pieces 45 sflow iptables Attack Detection Mitigation Database
?
46 ! --ip=1.2.3.4 example.com! ! --ip=1.2.3.4 example.com --qps=100! Reactive Rule
47 ! --ip=1.2.3.4 example.com --qps=500! ! example.com = FREE |
PAID! Reactive Rule ! --ip=1.2.3.4 example.com!
48 ! --ip=1.2.3.4 example.com --except www,n1,ns2 --qps=500! Reactive Rule !
example.com subdomains:! (www, ns1, ns2)! ! --ip=1.2.3.4 example.com! ! example.com = FREE | PAID!
49 Input Steam extra stream extra stream Output Stream Reactive
Rule
Chain of transformations 50 ! def dns_mitigation(attack, plan, subdomains):! domain
= attack['domain']! ! qps = 100! if plan[domain] == 'business':! qps = 500! ! mitigation =! attack['description'] + \! ' --qps=%s' % qps + \! ' --except=%s'.join(subdomains[domain])! ! return mitigation!
Fully composable 51
Putting it all together 52
Putting it all together 53 Mitigation Database sflow iptables Attack
Detection Reactive Automation 53
Gatebot: frequency 54 Gatebot actions per day 3 months
Gatebot: volume 55 1 week
Summary 56
The fight goes on 57 Malicious Attacker Internet Provider Origin
Server CloudFlare Server trust & safety team w orking w ith operators public outreach im proving our infrastructure
! ! • https://blog.cloudflare.com • https://github.com/cloudflare 58 marek@cloudflare.com @majek04 Thanks!
and good luck! @cfgatebot