SlideShare uma empresa Scribd logo
1 de 40
Baixar para ler offline
INTERNSHIP FINAL REPORT
Who are you?
Ritta Narita
(github:@naritta)
The University of Tokyo, Engineering M2	

Researched about Physic simulation
I’ve worked in some companies.	

2
Projects for Intern
hivemall: Original VM for Random Forests	

!
fluentd: Socket Manager with ServerEngine
3
hivemall: Original VM for Random Forests
4
What’s Random Forest ?
make many decision trees,	

accept a majority decision
decision tree(play golf or not)
to know the result of decision tree,	

need calculation for bound features.
humidity	

> 30 %?
whether 	

= sunny
wind speed 	

> 10 m/s ?
play golf
don’t play
don’t play
don’t play
yes
yes
yes
5
generate JS code	

→execute using eval	

!
At present, to calculate decision tree
if (x[0]==0){	

	

 if (x[1]>30){	

	

 	

 return 1;}	

	

 	

 ・・・	

}	

else {	

	

 return 1;	

}
x = [weather, humidity, wind]	

0=play golf, 1=don’t play
humidity	

> 30 %?
whether 	

= sunny
wind speed 	

> 10 m/s ?
play golf
don’t play
don’t play
don’t play
yes
yes
yes
6
!
!
due to using eval, can execute any code	

!
For example	

hostile JS code like infinite loop	

→burden for TD	

!
It’s difficult to restrict JS code	

→need restricted environment to calculate decision tree	

!
Problem for JS
7
Then
generate original op code from tree model	

→execute on originalVM
PUT x[1]	

PUT 0	

IFEQ 10	

!
・	

・	

・
x = [weather, humidity, wind]	

0=play golf, 1=don’t play
if (x[0]==0){	

	

 if (x[1]>30){	

	

 	

 return 1;}	

	

 	

 ・・・	

}	

else {	

	

 return 1;	

}
8
What’s the merit?
・can find illegal code like infinite loop easily	

・only for comparator, so very restricted	

・less op code, very fast
9
My work
op code featured for comparator	

 only PUSH, POP, GOTO, IF~	

!
can find infinite loop	

In this code, supposed not to have loop	

→don’t execute same code
10
hadoop version 2.6, Hive 1.2.0 (Tez 0.6.1)	

!
hadoop cluster size: c3.2xlarge 8 nodes	

!
!
randomforest	

!
 number of test examples in test_rf: 18083	

!
 number of trees: 500	

!
!
!
compile num: 500	

!
eval num: 500 * 18083	

!
Javascript : 1062.04 s
(Nashorn)	

!
VM: 106.84 s  
comparison with JS
10 times faster
11
Why don’t you use Java bytecode and ASM ?
12
Because of the number of class loading
for example, if every clients make 500 models…	

↓	

too many class loading
If using one class and 500 method, 	

It is same.
13
summary
・very restricted, can find illegal code	

!
・10 times faster	

!
・future prospects:	

	

 can make it even faster by binary code	

!
・merged in development branch	

	

 and will be released in v0.4	

14
fluentd: Socket Manager with ServerEngine
15
In fluentd v0.14
produce	

!
New multiprocess model
16
multiprocess at present
use in_multiprocess plugin	

have to use multi sockets and assign each ports by user
super	

visor
worker
worker
worker
port:	

24224
port:	

24226
port:	

24225
17
multiprocess v0.14
super	

visor
worker
worker
worker
port:	

24224
using Socket Manager, share one listening socket	

→can use multicore without any assignment
port:	

24224
port:	

24224
port:	

24224
Socket	

Manager	

server
Socket	

Manager	

client
Socket	

Manager	

client
Socket	

Manager	

client
18
in Windows
20
can use multicore power fully without unconsciousness	

setting file will get very simple 21
with SocketManagerwith in_multiprocess plugin
<source>	

type multiprocess	

<process>	

cmdline -c /etc/td-agent/td-agent-child1.conf	

</process>	

<process>	

cmdline -c /etc/td-agent/td-agent-child2.conf	

</process>	

</source>	

!
#/etc/td-agent/td-agent-child1.conf	

<source>	

type forward	

port 24224	

</source>	

!
#/etc/td-agent/td-agent-child2.conf	

<source>	

type forward	

port 24225	

</source>
<source>	

type forward	

port 24224	

</source>
setting when using 2 core
To implement Socket Manager, I used ServerEngine
worker
worker
worker
super	

visor
Server	

Engine
live restart
Heartbeat via pipe	

auto restart
22
ServerEngine is: a framework to implement 	

robust multiprocess servers like Unicorn.
Implementation (Unix)
②Unix Domain Socket (send_io file descriptor)
worker
worker
worker
Socket	

Manager	

client
Socket	

Manager	

client
Socket	

Manager	

client
FD
Spawn
Socket	

Manager	

server
super	

visor
Server	

Engine
①DRb (request listening socket)
24
Unix: very simple	

Windows: a little complex
main difference
1. can’t share socket by FD	

  in Windows, socket descriptor ≠ file descriptor	

  It doesn’t make sense to share FD	

  (have to use Winsock2 API to share sockets)	

!
2. have to lock accept	

	

 in unix, don’t need consider thundering herd	

 but do in windows.	

 
25
Implementation (Windows)
DRb
create socket from port and bind	

(WSASocket)	

↓	

duplicate exclusive socket by pid	

(WSADuplicateSocket)	

↓	

get socket protocol (WSAProcolInfo)
worker
worker
worker
Socket	

Manager	

server
Socket	

Manager	

client
Socket	

Manager	

client
Socket	

Manager	

client
from WSAProcolInfo,	

make WSASocket	

↓	

handle into FD	

↓	

IO.for_fd(FD)	

send this IO to Cool.io	

super	

visor
Server	

Engine
26
accept mutex
worker
worker
get 	

mutex
detach	

release 	

mutex
attach 	

listening socket	

to cool.io loop
accept
mutex
read and send data	

to buffer/output
server socket
get 	

mutex
detach	

release 	

mutex
attach 	

listening socket	

to cool.io loop
accept
read and send data	

to buffer/output
deal with post processing	

in this process as it is
other process can listen 	

while this process is dealing with data
27
rotation in order	

by accept mutex
 ①2376→②3456→③2696→④3388	

→①2376→②3456→ 28
As a result of test, 	

Thundering herd doesn’t occur in windows.	

Tentatively I implemented roughly with mutex,	

but I want to use IOCP like livuv in the future.	

!
Patches are welcome from Windows specialist!
29
benchmark result (unix)
AWS ubuntu 14.04 m4.xlarge
RPS IO
conventional
model
6798.69 	

/sec
1361.07 	

kb/s
new model	

(4 workers)
13743.02	

/sec
2751.29 	

kb/s
in_http → out_forward
30
benchmark result (windows)
AWS Microsoft Windows Server 2012 R2 m4.xlarge
RPS IO
conventional
model
1834.01	

/sec
385.07	

kb/s
new model	

(4 workers)
3513.31	

/sec
737.66	

kb/s
in_http → out_forward
31
Future work
・Buffering in multiprocess	

・accept mutex based IOCP…etc
summary
・Implemented fluentd Socket Manager with ServerEngine,	

and will be faster without consciousness.	

!
・There is details in ServerEngine Issue, 	

 you can test my forked branch(fluentd and ServerEngine)	

 and I’ll send PR after this report.
32
That’s all,Thank you!
33
appendix
Why don’t you use Object serialization?
35
Because of memory problem
When Random forests model is big and many customers use it, 	

It is too much memory consumption
36
ServerEngine is:
To implement Socket Manager, I used ServerEngine
a framework to implement 	

robust multiprocess servers like Unicorn.
37
how to use Socket Manager in fluentd side
!
#get socket manager	

socket_manager = ServerEngine::SocketManager.new_socket_manager 	

!
#get FD from socket manager	

fd = socket_manager.get_tcp(bind, port)	

!
#create listening socket from FD	

lsock = TCPServer.for_fd(fd.to_i)	

it doesn’t need consider about socket sharing in fluentd side,	

ServerEngine deal with it inside.
38
Benchmark Result
I’ll add multiprocess buffering function,	

After that I’ll do benchmark formally.	

!
Tentatively Show the rough result
40

Mais conteúdo relacionado

Mais procurados

nginx: writing your first module
nginx: writing your first modulenginx: writing your first module
nginx: writing your first module
redivy
 
Node js presentation
Node js presentationNode js presentation
Node js presentation
martincabrera
 
OpenNebula and SaltStack - OpenNebulaConf 2013
OpenNebula and SaltStack - OpenNebulaConf 2013OpenNebula and SaltStack - OpenNebulaConf 2013
OpenNebula and SaltStack - OpenNebulaConf 2013
databus.pro
 

Mais procurados (20)

Fluentd v1.0 in a nutshell
Fluentd v1.0 in a nutshellFluentd v1.0 in a nutshell
Fluentd v1.0 in a nutshell
 
Keynote - Fluentd meetup v14
Keynote - Fluentd meetup v14Keynote - Fluentd meetup v14
Keynote - Fluentd meetup v14
 
PyCon AU 2012 - Debugging Live Python Web Applications
PyCon AU 2012 - Debugging Live Python Web ApplicationsPyCon AU 2012 - Debugging Live Python Web Applications
PyCon AU 2012 - Debugging Live Python Web Applications
 
Fluentd v0.12 master guide
Fluentd v0.12 master guideFluentd v0.12 master guide
Fluentd v0.12 master guide
 
[231] the simplicity of cluster apps with circuit
[231] the simplicity of cluster apps with circuit[231] the simplicity of cluster apps with circuit
[231] the simplicity of cluster apps with circuit
 
Implementações paralelas
Implementações paralelasImplementações paralelas
Implementações paralelas
 
nginx: writing your first module
nginx: writing your first modulenginx: writing your first module
nginx: writing your first module
 
Ansible module development 101
Ansible module development 101Ansible module development 101
Ansible module development 101
 
Implement server push in flask framework
Implement server push in flask frameworkImplement server push in flask framework
Implement server push in flask framework
 
Bucks County Tech Meetup: node.js introduction
Bucks County Tech Meetup: node.js introductionBucks County Tech Meetup: node.js introduction
Bucks County Tech Meetup: node.js introduction
 
Introduction to node.js
Introduction to node.jsIntroduction to node.js
Introduction to node.js
 
Gude for C++11 in Apache Traffic Server
Gude for C++11 in Apache Traffic ServerGude for C++11 in Apache Traffic Server
Gude for C++11 in Apache Traffic Server
 
PyCon HK 2015 - Monitoring the performance of python web applications
PyCon HK 2015 -  Monitoring the performance of python web applicationsPyCon HK 2015 -  Monitoring the performance of python web applications
PyCon HK 2015 - Monitoring the performance of python web applications
 
NodeJs
NodeJsNodeJs
NodeJs
 
DevOps(4) : Ansible(2) - (MOSG)
DevOps(4) : Ansible(2) - (MOSG)DevOps(4) : Ansible(2) - (MOSG)
DevOps(4) : Ansible(2) - (MOSG)
 
Test driven infrastructure
Test driven infrastructureTest driven infrastructure
Test driven infrastructure
 
Python, do you even async?
Python, do you even async?Python, do you even async?
Python, do you even async?
 
Node js presentation
Node js presentationNode js presentation
Node js presentation
 
OpenNebula and SaltStack - OpenNebulaConf 2013
OpenNebula and SaltStack - OpenNebulaConf 2013OpenNebula and SaltStack - OpenNebulaConf 2013
OpenNebula and SaltStack - OpenNebulaConf 2013
 
Fluentd v1 and Roadmap
Fluentd v1 and RoadmapFluentd v1 and Roadmap
Fluentd v1 and Roadmap
 

Destaque

P2Pネットワークを利用した分散ファイルシステムの開発
P2Pネットワークを利用した分散ファイルシステムの開発P2Pネットワークを利用した分散ファイルシステムの開発
P2Pネットワークを利用した分散ファイルシステムの開発
shiftky
 
Testing Forest-Isomorphism in the Adjacency List Model
Testing Forest-Isomorphismin the Adjacency List ModelTesting Forest-Isomorphismin the Adjacency List Model
Testing Forest-Isomorphism in the Adjacency List Model
irrrrr
 
20 cv mil_models_for_words
20 cv mil_models_for_words20 cv mil_models_for_words
20 cv mil_models_for_words
zukun
 
Recommending Tags with a Model of Human Categorization
Recommending Tags with a Model of Human CategorizationRecommending Tags with a Model of Human Categorization
Recommending Tags with a Model of Human Categorization
Christoph Trattner
 
LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)
LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)
LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)
rchbeir
 

Destaque (20)

Treasure Data Summer Internship Final Report
Treasure Data Summer Internship Final ReportTreasure Data Summer Internship Final Report
Treasure Data Summer Internship Final Report
 
Internship final report@Treasure Data Inc.
Internship final report@Treasure Data Inc.Internship final report@Treasure Data Inc.
Internship final report@Treasure Data Inc.
 
並列データベースシステムの概念と原理
並列データベースシステムの概念と原理並列データベースシステムの概念と原理
並列データベースシステムの概念と原理
 
Embulk, an open-source plugin-based parallel bulk data loader
Embulk, an open-source plugin-based parallel bulk data loaderEmbulk, an open-source plugin-based parallel bulk data loader
Embulk, an open-source plugin-based parallel bulk data loader
 
P2Pネットワークを利用した分散ファイルシステムの開発
P2Pネットワークを利用した分散ファイルシステムの開発P2Pネットワークを利用した分散ファイルシステムの開発
P2Pネットワークを利用した分散ファイルシステムの開発
 
Fluentd meetup in japan
Fluentd meetup in japanFluentd meetup in japan
Fluentd meetup in japan
 
ICALP 2014 参加記
ICALP 2014 参加記ICALP 2014 参加記
ICALP 2014 参加記
 
Testing Forest-Isomorphism in the Adjacency List Model
Testing Forest-Isomorphismin the Adjacency List ModelTesting Forest-Isomorphismin the Adjacency List Model
Testing Forest-Isomorphism in the Adjacency List Model
 
トレジャーデータ 導入体験記 リブセンス編
トレジャーデータ 導入体験記 リブセンス編トレジャーデータ 導入体験記 リブセンス編
トレジャーデータ 導入体験記 リブセンス編
 
第2章アーキテクチャ
第2章アーキテクチャ第2章アーキテクチャ
第2章アーキテクチャ
 
EventSystemまわりの話@UnityFukuoka07
EventSystemまわりの話@UnityFukuoka07 EventSystemまわりの話@UnityFukuoka07
EventSystemまわりの話@UnityFukuoka07
 
tmu_science_cafe02
tmu_science_cafe02tmu_science_cafe02
tmu_science_cafe02
 
20 cv mil_models_for_words
20 cv mil_models_for_words20 cv mil_models_for_words
20 cv mil_models_for_words
 
Mathematical approach for Text Mining 1
Mathematical approach for Text Mining 1Mathematical approach for Text Mining 1
Mathematical approach for Text Mining 1
 
SpeakerLDA: Discovering Topics in Transcribed Multi-Speaker Audio Contents @ ...
SpeakerLDA: Discovering Topics in Transcribed Multi-Speaker Audio Contents @ ...SpeakerLDA: Discovering Topics in Transcribed Multi-Speaker Audio Contents @ ...
SpeakerLDA: Discovering Topics in Transcribed Multi-Speaker Audio Contents @ ...
 
Recommending Tags with a Model of Human Categorization
Recommending Tags with a Model of Human CategorizationRecommending Tags with a Model of Human Categorization
Recommending Tags with a Model of Human Categorization
 
Latent Semantic Indexing and Search Engines Optimimization (SEO)
Latent Semantic Indexing and Search Engines Optimimization (SEO)Latent Semantic Indexing and Search Engines Optimimization (SEO)
Latent Semantic Indexing and Search Engines Optimimization (SEO)
 
Analysis of Reviews on Sony Z3
Analysis of Reviews on Sony Z3Analysis of Reviews on Sony Z3
Analysis of Reviews on Sony Z3
 
LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)
LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)
LSI latent (par HATOUM Saria et DONGO ESCALANTE Irvin Franco)
 
Geometric Aspects of LSA
Geometric Aspects of LSAGeometric Aspects of LSA
Geometric Aspects of LSA
 

Semelhante a Treasure Data Summer Internship Final Report

Advanced off heap ipc
Advanced off heap ipcAdvanced off heap ipc
Advanced off heap ipc
Peter Lawrey
 
BP206 - Let's Give Your LotusScript a Tune-Up
BP206 - Let's Give Your LotusScript a Tune-Up BP206 - Let's Give Your LotusScript a Tune-Up
BP206 - Let's Give Your LotusScript a Tune-Up
Craig Schumann
 
Kettunen, miaubiz fuzzing at scale and in style
Kettunen, miaubiz   fuzzing at scale and in styleKettunen, miaubiz   fuzzing at scale and in style
Kettunen, miaubiz fuzzing at scale and in style
DefconRussia
 
44 con slides
44 con slides44 con slides
44 con slides
geeksec80
 
44 con slides (1)
44 con slides (1)44 con slides (1)
44 con slides (1)
geeksec80
 
Scalable Socket Server by Aryo
Scalable Socket Server by AryoScalable Socket Server by Aryo
Scalable Socket Server by Aryo
Agate Studio
 
Efficient use of NodeJS
Efficient use of NodeJSEfficient use of NodeJS
Efficient use of NodeJS
Yura Bogdanov
 

Semelhante a Treasure Data Summer Internship Final Report (20)

Advanced off heap ipc
Advanced off heap ipcAdvanced off heap ipc
Advanced off heap ipc
 
NodeJS : Communication and Round Robin Way
NodeJS : Communication and Round Robin WayNodeJS : Communication and Round Robin Way
NodeJS : Communication and Round Robin Way
 
BP206 - Let's Give Your LotusScript a Tune-Up
BP206 - Let's Give Your LotusScript a Tune-Up BP206 - Let's Give Your LotusScript a Tune-Up
BP206 - Let's Give Your LotusScript a Tune-Up
 
Objects? No thanks!
Objects? No thanks!Objects? No thanks!
Objects? No thanks!
 
Distributed systems at ok.ru #rigadevday
Distributed systems at ok.ru #rigadevdayDistributed systems at ok.ru #rigadevday
Distributed systems at ok.ru #rigadevday
 
Nodejs
NodejsNodejs
Nodejs
 
The why and how of moving to PHP 5.5/5.6
The why and how of moving to PHP 5.5/5.6The why and how of moving to PHP 5.5/5.6
The why and how of moving to PHP 5.5/5.6
 
Reusing your existing software on Android
Reusing your existing software on AndroidReusing your existing software on Android
Reusing your existing software on Android
 
Kettunen, miaubiz fuzzing at scale and in style
Kettunen, miaubiz   fuzzing at scale and in styleKettunen, miaubiz   fuzzing at scale and in style
Kettunen, miaubiz fuzzing at scale and in style
 
Better Code: Concurrency
Better Code: ConcurrencyBetter Code: Concurrency
Better Code: Concurrency
 
Introduction to Node.js
Introduction to Node.jsIntroduction to Node.js
Introduction to Node.js
 
Oleksandr Smoktal "Parallel Seismic Data Processing Using OpenMP"
Oleksandr Smoktal "Parallel Seismic Data Processing Using OpenMP"Oleksandr Smoktal "Parallel Seismic Data Processing Using OpenMP"
Oleksandr Smoktal "Parallel Seismic Data Processing Using OpenMP"
 
44 con slides
44 con slides44 con slides
44 con slides
 
44 con slides (1)
44 con slides (1)44 con slides (1)
44 con slides (1)
 
Multi-core Node.pdf
Multi-core Node.pdfMulti-core Node.pdf
Multi-core Node.pdf
 
Scalable Socket Server by Aryo
Scalable Socket Server by AryoScalable Socket Server by Aryo
Scalable Socket Server by Aryo
 
How to make a large C++-code base manageable
How to make a large C++-code base manageableHow to make a large C++-code base manageable
How to make a large C++-code base manageable
 
StudioSL Presentation in Grenoble 2011
StudioSL Presentation in Grenoble 2011StudioSL Presentation in Grenoble 2011
StudioSL Presentation in Grenoble 2011
 
Virtualization and Socket Programing
Virtualization and Socket ProgramingVirtualization and Socket Programing
Virtualization and Socket Programing
 
Efficient use of NodeJS
Efficient use of NodeJSEfficient use of NodeJS
Efficient use of NodeJS
 

Último

DeepFakes presentation : brief idea of DeepFakes
DeepFakes presentation : brief idea of DeepFakesDeepFakes presentation : brief idea of DeepFakes
DeepFakes presentation : brief idea of DeepFakes
MayuraD1
 
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak HamilCara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Kandungan 087776558899
 
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
AldoGarca30
 
Hospital management system project report.pdf
Hospital management system project report.pdfHospital management system project report.pdf
Hospital management system project report.pdf
Kamal Acharya
 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
ssuser89054b
 
Kuwait City MTP kit ((+919101817206)) Buy Abortion Pills Kuwait
Kuwait City MTP kit ((+919101817206)) Buy Abortion Pills KuwaitKuwait City MTP kit ((+919101817206)) Buy Abortion Pills Kuwait
Kuwait City MTP kit ((+919101817206)) Buy Abortion Pills Kuwait
jaanualu31
 

Último (20)

kiln thermal load.pptx kiln tgermal load
kiln thermal load.pptx kiln tgermal loadkiln thermal load.pptx kiln tgermal load
kiln thermal load.pptx kiln tgermal load
 
Double Revolving field theory-how the rotor develops torque
Double Revolving field theory-how the rotor develops torqueDouble Revolving field theory-how the rotor develops torque
Double Revolving field theory-how the rotor develops torque
 
AIRCANVAS[1].pdf mini project for btech students
AIRCANVAS[1].pdf mini project for btech studentsAIRCANVAS[1].pdf mini project for btech students
AIRCANVAS[1].pdf mini project for btech students
 
DeepFakes presentation : brief idea of DeepFakes
DeepFakes presentation : brief idea of DeepFakesDeepFakes presentation : brief idea of DeepFakes
DeepFakes presentation : brief idea of DeepFakes
 
Block diagram reduction techniques in control systems.ppt
Block diagram reduction techniques in control systems.pptBlock diagram reduction techniques in control systems.ppt
Block diagram reduction techniques in control systems.ppt
 
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak HamilCara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
 
Engineering Drawing focus on projection of planes
Engineering Drawing focus on projection of planesEngineering Drawing focus on projection of planes
Engineering Drawing focus on projection of planes
 
Employee leave management system project.
Employee leave management system project.Employee leave management system project.
Employee leave management system project.
 
GEAR TRAIN- BASIC CONCEPTS AND WORKING PRINCIPLE
GEAR TRAIN- BASIC CONCEPTS AND WORKING PRINCIPLEGEAR TRAIN- BASIC CONCEPTS AND WORKING PRINCIPLE
GEAR TRAIN- BASIC CONCEPTS AND WORKING PRINCIPLE
 
Online food ordering system project report.pdf
Online food ordering system project report.pdfOnline food ordering system project report.pdf
Online food ordering system project report.pdf
 
data_management_and _data_science_cheat_sheet.pdf
data_management_and _data_science_cheat_sheet.pdfdata_management_and _data_science_cheat_sheet.pdf
data_management_and _data_science_cheat_sheet.pdf
 
A CASE STUDY ON CERAMIC INDUSTRY OF BANGLADESH.pptx
A CASE STUDY ON CERAMIC INDUSTRY OF BANGLADESH.pptxA CASE STUDY ON CERAMIC INDUSTRY OF BANGLADESH.pptx
A CASE STUDY ON CERAMIC INDUSTRY OF BANGLADESH.pptx
 
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
 
Unleashing the Power of the SORA AI lastest leap
Unleashing the Power of the SORA AI lastest leapUnleashing the Power of the SORA AI lastest leap
Unleashing the Power of the SORA AI lastest leap
 
Tamil Call Girls Bhayandar WhatsApp +91-9930687706, Best Service
Tamil Call Girls Bhayandar WhatsApp +91-9930687706, Best ServiceTamil Call Girls Bhayandar WhatsApp +91-9930687706, Best Service
Tamil Call Girls Bhayandar WhatsApp +91-9930687706, Best Service
 
Computer Networks Basics of Network Devices
Computer Networks  Basics of Network DevicesComputer Networks  Basics of Network Devices
Computer Networks Basics of Network Devices
 
Hospital management system project report.pdf
Hospital management system project report.pdfHospital management system project report.pdf
Hospital management system project report.pdf
 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
 
DC MACHINE-Motoring and generation, Armature circuit equation
DC MACHINE-Motoring and generation, Armature circuit equationDC MACHINE-Motoring and generation, Armature circuit equation
DC MACHINE-Motoring and generation, Armature circuit equation
 
Kuwait City MTP kit ((+919101817206)) Buy Abortion Pills Kuwait
Kuwait City MTP kit ((+919101817206)) Buy Abortion Pills KuwaitKuwait City MTP kit ((+919101817206)) Buy Abortion Pills Kuwait
Kuwait City MTP kit ((+919101817206)) Buy Abortion Pills Kuwait
 

Treasure Data Summer Internship Final Report

  • 2. Who are you? Ritta Narita (github:@naritta) The University of Tokyo, Engineering M2 Researched about Physic simulation I’ve worked in some companies. 2
  • 3. Projects for Intern hivemall: Original VM for Random Forests ! fluentd: Socket Manager with ServerEngine 3
  • 4. hivemall: Original VM for Random Forests 4
  • 5. What’s Random Forest ? make many decision trees, accept a majority decision decision tree(play golf or not) to know the result of decision tree, need calculation for bound features. humidity > 30 %? whether = sunny wind speed > 10 m/s ? play golf don’t play don’t play don’t play yes yes yes 5
  • 6. generate JS code →execute using eval ! At present, to calculate decision tree if (x[0]==0){ if (x[1]>30){ return 1;} ・・・ } else { return 1; } x = [weather, humidity, wind] 0=play golf, 1=don’t play humidity > 30 %? whether = sunny wind speed > 10 m/s ? play golf don’t play don’t play don’t play yes yes yes 6
  • 7. ! ! due to using eval, can execute any code ! For example hostile JS code like infinite loop →burden for TD ! It’s difficult to restrict JS code →need restricted environment to calculate decision tree ! Problem for JS 7
  • 8. Then generate original op code from tree model →execute on originalVM PUT x[1] PUT 0 IFEQ 10 ! ・ ・ ・ x = [weather, humidity, wind] 0=play golf, 1=don’t play if (x[0]==0){ if (x[1]>30){ return 1;} ・・・ } else { return 1; } 8
  • 9. What’s the merit? ・can find illegal code like infinite loop easily ・only for comparator, so very restricted ・less op code, very fast 9
  • 10. My work op code featured for comparator  only PUSH, POP, GOTO, IF~ ! can find infinite loop In this code, supposed not to have loop →don’t execute same code 10
  • 11. hadoop version 2.6, Hive 1.2.0 (Tez 0.6.1) ! hadoop cluster size: c3.2xlarge 8 nodes ! ! randomforest !  number of test examples in test_rf: 18083 !  number of trees: 500 ! ! ! compile num: 500 ! eval num: 500 * 18083 ! Javascript : 1062.04 s (Nashorn) ! VM: 106.84 s   comparison with JS 10 times faster 11
  • 12. Why don’t you use Java bytecode and ASM ? 12
  • 13. Because of the number of class loading for example, if every clients make 500 models… ↓ too many class loading If using one class and 500 method, It is same. 13
  • 14. summary ・very restricted, can find illegal code ! ・10 times faster ! ・future prospects: can make it even faster by binary code ! ・merged in development branch and will be released in v0.4 14
  • 15. fluentd: Socket Manager with ServerEngine 15
  • 16. In fluentd v0.14 produce ! New multiprocess model 16
  • 17. multiprocess at present use in_multiprocess plugin have to use multi sockets and assign each ports by user super visor worker worker worker port: 24224 port: 24226 port: 24225 17
  • 18. multiprocess v0.14 super visor worker worker worker port: 24224 using Socket Manager, share one listening socket →can use multicore without any assignment port: 24224 port: 24224 port: 24224 Socket Manager server Socket Manager client Socket Manager client Socket Manager client 18
  • 19.
  • 21. can use multicore power fully without unconsciousness setting file will get very simple 21 with SocketManagerwith in_multiprocess plugin <source> type multiprocess <process> cmdline -c /etc/td-agent/td-agent-child1.conf </process> <process> cmdline -c /etc/td-agent/td-agent-child2.conf </process> </source> ! #/etc/td-agent/td-agent-child1.conf <source> type forward port 24224 </source> ! #/etc/td-agent/td-agent-child2.conf <source> type forward port 24225 </source> <source> type forward port 24224 </source> setting when using 2 core
  • 22. To implement Socket Manager, I used ServerEngine worker worker worker super visor Server Engine live restart Heartbeat via pipe auto restart 22 ServerEngine is: a framework to implement robust multiprocess servers like Unicorn.
  • 23.
  • 24. Implementation (Unix) ②Unix Domain Socket (send_io file descriptor) worker worker worker Socket Manager client Socket Manager client Socket Manager client FD Spawn Socket Manager server super visor Server Engine ①DRb (request listening socket) 24
  • 25. Unix: very simple Windows: a little complex main difference 1. can’t share socket by FD   in Windows, socket descriptor ≠ file descriptor   It doesn’t make sense to share FD   (have to use Winsock2 API to share sockets) ! 2. have to lock accept in unix, don’t need consider thundering herd  but do in windows.   25
  • 26. Implementation (Windows) DRb create socket from port and bind (WSASocket) ↓ duplicate exclusive socket by pid (WSADuplicateSocket) ↓ get socket protocol (WSAProcolInfo) worker worker worker Socket Manager server Socket Manager client Socket Manager client Socket Manager client from WSAProcolInfo, make WSASocket ↓ handle into FD ↓ IO.for_fd(FD) send this IO to Cool.io super visor Server Engine 26
  • 27. accept mutex worker worker get mutex detach release mutex attach listening socket to cool.io loop accept mutex read and send data to buffer/output server socket get mutex detach release mutex attach listening socket to cool.io loop accept read and send data to buffer/output deal with post processing in this process as it is other process can listen while this process is dealing with data 27
  • 28. rotation in order by accept mutex  ①2376→②3456→③2696→④3388 →①2376→②3456→ 28
  • 29. As a result of test, Thundering herd doesn’t occur in windows. Tentatively I implemented roughly with mutex, but I want to use IOCP like livuv in the future. ! Patches are welcome from Windows specialist! 29
  • 30. benchmark result (unix) AWS ubuntu 14.04 m4.xlarge RPS IO conventional model 6798.69 /sec 1361.07 kb/s new model (4 workers) 13743.02 /sec 2751.29 kb/s in_http → out_forward 30
  • 31. benchmark result (windows) AWS Microsoft Windows Server 2012 R2 m4.xlarge RPS IO conventional model 1834.01 /sec 385.07 kb/s new model (4 workers) 3513.31 /sec 737.66 kb/s in_http → out_forward 31
  • 32. Future work ・Buffering in multiprocess ・accept mutex based IOCP…etc summary ・Implemented fluentd Socket Manager with ServerEngine, and will be faster without consciousness. ! ・There is details in ServerEngine Issue,  you can test my forked branch(fluentd and ServerEngine)  and I’ll send PR after this report. 32
  • 35. Why don’t you use Object serialization? 35
  • 36. Because of memory problem When Random forests model is big and many customers use it, It is too much memory consumption 36
  • 37. ServerEngine is: To implement Socket Manager, I used ServerEngine a framework to implement robust multiprocess servers like Unicorn. 37
  • 38. how to use Socket Manager in fluentd side ! #get socket manager socket_manager = ServerEngine::SocketManager.new_socket_manager ! #get FD from socket manager fd = socket_manager.get_tcp(bind, port) ! #create listening socket from FD lsock = TCPServer.for_fd(fd.to_i) it doesn’t need consider about socket sharing in fluentd side, ServerEngine deal with it inside. 38
  • 39.
  • 40. Benchmark Result I’ll add multiprocess buffering function, After that I’ll do benchmark formally. ! Tentatively Show the rough result 40