SlideShare une entreprise Scribd logo
1  sur  27
Improve mobile application performance
Mobility for Business – Oct 2014
2
Why is speed important ?
The impact of
mobility
How to test ?
Yesterdays
solutions
Todays
solutions
3
Our brain
Sources : Public Library of Science, Appdynamics, “Timing responses to questions in dialogue”, Sofia Strömbergsson, Anna
Hjalmarsson, Jens Edlund, David House - Department of Speech, Music and Hearing, KTH, Stockholm, Sweden
Arabs numbers(150ms)
Comparison (190ms)
Movement (330ms)
Error corections (470ms)
Reaction time
100 to 500 ms
Response time
100 to 500 ms
Memory
Sensory memory
Duration: 0,5 max for visual
2 sec per audio
Short term memory
Duration : 30 seconds max
Long term memory
Average response time to
a question (ms)
460
310
180
90
Question ouverte
Question à réponse
simple
Question dont la
réponse est Oui/Non
Question incluant les
réponses possibles
4
Effect of reaction time of an application
on humans
4
Reaction time
of an
application
Impact /
sensation on
the user
0,1 Seconds0,5 2 3 5 101 4
« Instant »
Perfect (eq.
Human
Realtionship)
« normal »
Time
« slow »
40 to 60% drops
on the web
desktop
« very slow »
30 to 40% drops
on the web
desktop
Loss of attention,
nervousness,
tiredness
No waiting
55
0,1 Seconds0,5 2 3 5 101 4
« Instant »
Temps
« normal »
Average time
Web desktop
Average time
Web mobile
Impact /
sensation on
the user
Reaction time
of an
application
Effect of reaction time of an application
on humans
Perfect (eq.
Human
Realtionship)
« slow »
40 to 60% drops
on the web
desktop
No waiting
« very slow »
30 to 40% drops
on the web
desktop
Loss of attention,
nervousness,
tiredness
66
0,1 Seconds0,5 2 3 5 101 4
« Instant »
Temps
« normal »
Average time
Web desktop
Average time
Web mobile
Impact /
sensation on
the user
Reaction time
of an
application
Effect of reaction time of an application
On humans
Perfect (eq.
Human
Realtionship)
« slow »
40 to 60% drops
on the web
desktop
No waiting
« very slow »
30 to 40% drops
on the web
desktop
Loss of attention
feeling nervous
tired
Revenue Decline Decline in Adoption
Decreased
productivity
Consequences
Amazon: If the
display time
increases by 100 ms,
the revenue goes
down to 1%
Google : If the
display time
increases 500ms ,
research fell by 25%
Compared to a site
loading in 1 second , a
site loading in 3
seconds losses 22%
of page views
,increase rebounds by
50 % and generate
22% less
transformations
A slow website requires
50% concentration of
additional efforts
78 % of users feel
stressed or angry when
using a slow website
Increase in usage time
for employees
Sources :Jakob Nielsen : http://www.useit.com/alertbox/response-times.html, Foviance: http://www.foviance.com/what-we-think/wen-
stress-a-wake-up-call-for european-business/, Harris interactive: http://www.tealeaf.com/news/news-releases/2011/tealeaf-announces-
new-mobile-research.php, PhoCusWright : http://www.phocuswright.com/free_reports/consumer-response-to-travel-site-performance,
Redhat http://www.webperformancetoday.com/2010/07/01/the-best-graphs-of-velocity/
7
Why is performance
important ?
The Impact of mobility
How to test ?
Yesterdays
solutions
Todays
Solutions
8
The sources of performance
Page
construction
Answers
Request
Display Treatment
Performance
=
Device processing
+
Time to communication
+
Server processing
Latency
Debit
Amount of Data
Quantity of requests
Protocol
9
TCP SYN « Hi !»
TCP SYN ACK « Hi ?! »
TCP ACK « There you are ! »
HTTP GET « I want this »
HTTP 200« OK »
[index.html] « Here it is : ….»
Note: Secured version (SSL ), the dialogue before data recovery requires a turn and return to give
The " encryption keys "
HTTP
The most common protocol (HTTP)
= 2 round trips
Time for one round trip = latency
Estimated latency of the different
types of network
10Source : http://www.digitalsociety.org/2010/08/conflating-broadband-speed-with-internet-speed-is-misleading/
http://www.dslreports.com/
Because 2 round trips are required by HTTP, first « useful » data reaches
application in 40ms for ADSL , in 300ms in 3G
0 100 200 300 400
EDGE/2G
3G
Optical fiber inter-continental
Optical fiber inter-country
ADSL
Private local Network
11
4G delivers us ?
11http://opensignal.com/reports/state-of-lte-q1-2014/
Yes, it multiple the
download by 4
More dedicated bandwidth depends on the overall traffic and
mobile traffic increased by 68 % in France
But it reduces latency
just by 40 %
12
Why is performance
important ?
The Impact of
mobility
How to test ?
Yesterdays
solutions ?
Todays solutions
?
13
Tools to analyse application performance
(APM : Application Performance Management)
Synthetic testing Real User Monitoring (RUM)
Probe to integrate into applications or
based on emulations
Objectives : crash, display time , ...
by screens / by request
Actors: Compuware Riverbet , AppDynamics , New Relic , Akamai (
Movbitest ) , Keynote Systems, Witbe, Dynatrace, Aternity, Newrelic, …
Robots testing applications on
devices
14
Why is performance
important ?
The impact of
mobility
How to test ?
Yesterdays solution
Todays solutions
15
Time to communication
is a function of
Perception,
latency,
debit,
Amount of data,
Number of requests,
Protocol.
16
Display items quickly
Programmation : flush , Post-load , pre-load
CDN
Akamai Technologies, Edgecast,
level 3, Amazon, OVH, CDNetworks
Cache memories
At the server level and devices
Prepare ahead cache
Infrastructures
Bandwidth & CPU
Compress
Gzip/deflate via accept-encoding
/ content-encoding
Data formats
Ex : JSON > HTML
Binary communication protocol
To reduce the amount of data to be transmitted
Programming Quality
Reduce the number of request by grouping , Putting scripts after the
maximum requests , Avoid interdependent queries, Web: avoid CSS
expressions, Redirects , ... .
Time to communication
is a function of
Perception,
latency,
debit,
Amount of data,
Number of requests,
Protocol.
17
Why is performance
important ?
The impact of
mobility
How to test ?
Yesterdays
solutions
Todays solutions
18
Latency
10 to 20ms
Latency
80 to 160 ms
Source of the mobility problem
19
19
1st Goal: Reduce the
amount of data to be
transferred to the mobile
2nd goal: To reduce the
number of round trips for
the same amount of data
Source of the mobility problem
Latency
10 to 20ms
Latency
80 to 160 ms
20
Sources
Server
« middleware mobile »
Agregation
Cache Cache
Orchestration
Eliminate
unnecessary data
Avoid the multiple
calls from the mobile
(*)
Send only new data
Using bidirectional
protocol
(*) Especially the application requires interdependent calls ( if the value of A to be found in such a server is X, then search B in another server)
21
« Hello !»
« Yes ?! »
« Cool ! »
« I want this »
« OK »
« Here! »
HTTP
Page 1
Page 2
« Hello !»
« Again?! »
« Cool ! »
« I want this »
« OK »
« Here! »
Page 1 – Data refresh (*)
« Yeah … »
« Cool !»
« OK »
« Here! »
« Hello !»
« Yes ?! »
« I want this »
« OK »
« Tiens! »
Page 1
2nd page
«I want this »
« Here! »
1st page – data refresh
« Here! »
Bidirectional protocol
Note: (* ) upgradeable scenario with long polling
Compared to the conventional protocol and
two-way protocol
« Hello !»
« I want this »
« Cool ? »
22
Transferred data for the same transaction
Mobile application for trading
Transferred data for the same transaction
Mobile application for sports betting
298
95
1.6
0
50
100
150
200
250
300
350
Web app Mobile app Motwin app
Ko Ko
608
243
48
0
100
200
300
400
500
600
700
Web app Mobile app Motwin app
-99% -92%
The impact of these new solutions on the
quantity of transferred data
23
Do It Yourself or BUY/RENT
Motwin Platform
24
Implementation of the Motwin platform
25
3 sensitive segments performance
Trading & Betting
• Access to real-time data
• UI faster for customers in a hurry to buy / sell / bet
Collaborators Tools
• Real time CRM/ERP
• Assisted maintenance
IoT real time BtB
• Configuration, Maintenance and Monitoring
• IU efficient and real - time control objects
25
Critical data
Efficiency and Productivity
Critical systems
26
Reactivity Disponibility Speed
deployment
Maintenance Cost
and Scalability
-30 to 70% 90% less data
Management
Microcuts
x 7
Beyond the
first OS
-30% to -60%
Thank You !
Stand : E32B
27
community@motwin.com
www.motwin.com
@motwin_inc

Contenu connexe

Similaire à 20141007 latency prez - eng

PROGNOSIS for Web Applications
PROGNOSIS for Web ApplicationsPROGNOSIS for Web Applications
PROGNOSIS for Web Applicationsantoni0martin
 
Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...
Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...
Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...Mia Yuan Cao
 
Spicing Up Your Vanilla With a Custom Look and Feel
Spicing Up Your Vanilla With a Custom Look and FeelSpicing Up Your Vanilla With a Custom Look and Feel
Spicing Up Your Vanilla With a Custom Look and Feeljpberetz
 
Show Me the Money: Connecting Performance Engineering to Real Business Results
Show Me the Money: Connecting Performance Engineering to Real Business ResultsShow Me the Money: Connecting Performance Engineering to Real Business Results
Show Me the Money: Connecting Performance Engineering to Real Business ResultsCorrelsense
 
Performance Testing For Small Projects
Performance Testing For Small ProjectsPerformance Testing For Small Projects
Performance Testing For Small ProjectsMaksym Voitko
 
RTI Data-Distribution Service (DDS) Master Class 2011
RTI Data-Distribution Service (DDS) Master Class 2011RTI Data-Distribution Service (DDS) Master Class 2011
RTI Data-Distribution Service (DDS) Master Class 2011Gerardo Pardo-Castellote
 
MeasureWorks - Social Mentions as a Performance KPI
MeasureWorks - Social Mentions as a Performance KPIMeasureWorks - Social Mentions as a Performance KPI
MeasureWorks - Social Mentions as a Performance KPIMeasureWorks
 
MeasureWorks - eCommerce Live 2017 - Designing for Time & Conversion
MeasureWorks - eCommerce Live 2017 - Designing for Time & ConversionMeasureWorks - eCommerce Live 2017 - Designing for Time & Conversion
MeasureWorks - eCommerce Live 2017 - Designing for Time & ConversionMeasureWorks
 
MeasureWorks - eCommerce Live - Designing Time & Conversion
MeasureWorks -  eCommerce Live - Designing Time & ConversionMeasureWorks -  eCommerce Live - Designing Time & Conversion
MeasureWorks - eCommerce Live - Designing Time & Conversionnlwebperf
 
How to test performance and not die trying
How to test performance and not die tryingHow to test performance and not die trying
How to test performance and not die tryingDiego Cardozo
 
Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...
Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...
Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...WSO2
 
Applying a Methodical Approach to Website Performance
Applying a Methodical Approach to Website PerformanceApplying a Methodical Approach to Website Performance
Applying a Methodical Approach to Website PerformancePostSharp Technologies
 
Web Performance BootCamp 2013
Web Performance BootCamp 2013Web Performance BootCamp 2013
Web Performance BootCamp 2013Daniel Austin
 
MeasureWorks - Design for Fast Experiences (Startup session).key
MeasureWorks  - Design for Fast Experiences (Startup session).keyMeasureWorks  - Design for Fast Experiences (Startup session).key
MeasureWorks - Design for Fast Experiences (Startup session).keyMeasureWorks
 
A Beginner's Guide to Application Load Testing
A Beginner's Guide to Application Load TestingA Beginner's Guide to Application Load Testing
A Beginner's Guide to Application Load TestingBirgit Pauli-Haack
 
MeasureWorks - Design for Fast Experiences
MeasureWorks - Design for Fast ExperiencesMeasureWorks - Design for Fast Experiences
MeasureWorks - Design for Fast ExperiencesMeasureWorks
 
Mobile presentation - Sydney Online Retailer - 26 Sep 2011
Mobile presentation - Sydney Online Retailer - 26 Sep 2011Mobile presentation - Sydney Online Retailer - 26 Sep 2011
Mobile presentation - Sydney Online Retailer - 26 Sep 2011Craig Sullivan
 
Dg park introduction
Dg park introductionDg park introduction
Dg park introductionDeokgun Park
 

Similaire à 20141007 latency prez - eng (20)

PROGNOSIS for Web Applications
PROGNOSIS for Web ApplicationsPROGNOSIS for Web Applications
PROGNOSIS for Web Applications
 
Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...
Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...
Real-Time Analytics and Visualization of Streaming Big Data with JReport & Sc...
 
Spicing Up Your Vanilla With a Custom Look and Feel
Spicing Up Your Vanilla With a Custom Look and FeelSpicing Up Your Vanilla With a Custom Look and Feel
Spicing Up Your Vanilla With a Custom Look and Feel
 
Modern Web Applications
Modern Web ApplicationsModern Web Applications
Modern Web Applications
 
Show Me the Money: Connecting Performance Engineering to Real Business Results
Show Me the Money: Connecting Performance Engineering to Real Business ResultsShow Me the Money: Connecting Performance Engineering to Real Business Results
Show Me the Money: Connecting Performance Engineering to Real Business Results
 
Performance Testing For Small Projects
Performance Testing For Small ProjectsPerformance Testing For Small Projects
Performance Testing For Small Projects
 
RTI Data-Distribution Service (DDS) Master Class 2011
RTI Data-Distribution Service (DDS) Master Class 2011RTI Data-Distribution Service (DDS) Master Class 2011
RTI Data-Distribution Service (DDS) Master Class 2011
 
MeasureWorks - Social Mentions as a Performance KPI
MeasureWorks - Social Mentions as a Performance KPIMeasureWorks - Social Mentions as a Performance KPI
MeasureWorks - Social Mentions as a Performance KPI
 
MeasureWorks - eCommerce Live 2017 - Designing for Time & Conversion
MeasureWorks - eCommerce Live 2017 - Designing for Time & ConversionMeasureWorks - eCommerce Live 2017 - Designing for Time & Conversion
MeasureWorks - eCommerce Live 2017 - Designing for Time & Conversion
 
MeasureWorks - eCommerce Live - Designing Time & Conversion
MeasureWorks -  eCommerce Live - Designing Time & ConversionMeasureWorks -  eCommerce Live - Designing Time & Conversion
MeasureWorks - eCommerce Live - Designing Time & Conversion
 
How to test performance and not die trying
How to test performance and not die tryingHow to test performance and not die trying
How to test performance and not die trying
 
Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...
Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...
Introduction to Big Data Analytics: Batch, Real-Time, and the Best of Both Wo...
 
Watching websites
Watching websitesWatching websites
Watching websites
 
Applying a Methodical Approach to Website Performance
Applying a Methodical Approach to Website PerformanceApplying a Methodical Approach to Website Performance
Applying a Methodical Approach to Website Performance
 
Web Performance BootCamp 2013
Web Performance BootCamp 2013Web Performance BootCamp 2013
Web Performance BootCamp 2013
 
MeasureWorks - Design for Fast Experiences (Startup session).key
MeasureWorks  - Design for Fast Experiences (Startup session).keyMeasureWorks  - Design for Fast Experiences (Startup session).key
MeasureWorks - Design for Fast Experiences (Startup session).key
 
A Beginner's Guide to Application Load Testing
A Beginner's Guide to Application Load TestingA Beginner's Guide to Application Load Testing
A Beginner's Guide to Application Load Testing
 
MeasureWorks - Design for Fast Experiences
MeasureWorks - Design for Fast ExperiencesMeasureWorks - Design for Fast Experiences
MeasureWorks - Design for Fast Experiences
 
Mobile presentation - Sydney Online Retailer - 26 Sep 2011
Mobile presentation - Sydney Online Retailer - 26 Sep 2011Mobile presentation - Sydney Online Retailer - 26 Sep 2011
Mobile presentation - Sydney Online Retailer - 26 Sep 2011
 
Dg park introduction
Dg park introductionDg park introduction
Dg park introduction
 

Dernier

Search Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfSearch Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfRankYa
 
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyCommit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyAlfredo García Lavilla
 
Training state-of-the-art general text embedding
Training state-of-the-art general text embeddingTraining state-of-the-art general text embedding
Training state-of-the-art general text embeddingZilliz
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsMemoori
 
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024BookNet Canada
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr LapshynFwdays
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationSafe Software
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxhariprasad279825
 
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...Fwdays
 
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLDeveloper Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLScyllaDB
 
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr BaganFwdays
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfAlex Barbosa Coqueiro
 
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platformsDevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platformsSergiu Bodiu
 
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostLeverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostZilliz
 
My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024The Digital Insurer
 
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024BookNet Canada
 
Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 3652toLead Limited
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Enterprise Knowledge
 
Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!Commit University
 

Dernier (20)

Search Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfSearch Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdf
 
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyCommit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easy
 
Training state-of-the-art general text embedding
Training state-of-the-art general text embeddingTraining state-of-the-art general text embedding
Training state-of-the-art general text embedding
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
 
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
 
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptxE-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptx
 
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
 
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLDeveloper Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQL
 
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdf
 
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platformsDevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platforms
 
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostLeverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
 
My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024
 
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
 
Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024
 
Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!
 

20141007 latency prez - eng

  • 1. Improve mobile application performance Mobility for Business – Oct 2014
  • 2. 2 Why is speed important ? The impact of mobility How to test ? Yesterdays solutions Todays solutions
  • 3. 3 Our brain Sources : Public Library of Science, Appdynamics, “Timing responses to questions in dialogue”, Sofia Strömbergsson, Anna Hjalmarsson, Jens Edlund, David House - Department of Speech, Music and Hearing, KTH, Stockholm, Sweden Arabs numbers(150ms) Comparison (190ms) Movement (330ms) Error corections (470ms) Reaction time 100 to 500 ms Response time 100 to 500 ms Memory Sensory memory Duration: 0,5 max for visual 2 sec per audio Short term memory Duration : 30 seconds max Long term memory Average response time to a question (ms) 460 310 180 90 Question ouverte Question à réponse simple Question dont la réponse est Oui/Non Question incluant les réponses possibles
  • 4. 4 Effect of reaction time of an application on humans 4 Reaction time of an application Impact / sensation on the user 0,1 Seconds0,5 2 3 5 101 4 « Instant » Perfect (eq. Human Realtionship) « normal » Time « slow » 40 to 60% drops on the web desktop « very slow » 30 to 40% drops on the web desktop Loss of attention, nervousness, tiredness No waiting
  • 5. 55 0,1 Seconds0,5 2 3 5 101 4 « Instant » Temps « normal » Average time Web desktop Average time Web mobile Impact / sensation on the user Reaction time of an application Effect of reaction time of an application on humans Perfect (eq. Human Realtionship) « slow » 40 to 60% drops on the web desktop No waiting « very slow » 30 to 40% drops on the web desktop Loss of attention, nervousness, tiredness
  • 6. 66 0,1 Seconds0,5 2 3 5 101 4 « Instant » Temps « normal » Average time Web desktop Average time Web mobile Impact / sensation on the user Reaction time of an application Effect of reaction time of an application On humans Perfect (eq. Human Realtionship) « slow » 40 to 60% drops on the web desktop No waiting « very slow » 30 to 40% drops on the web desktop Loss of attention feeling nervous tired Revenue Decline Decline in Adoption Decreased productivity Consequences Amazon: If the display time increases by 100 ms, the revenue goes down to 1% Google : If the display time increases 500ms , research fell by 25% Compared to a site loading in 1 second , a site loading in 3 seconds losses 22% of page views ,increase rebounds by 50 % and generate 22% less transformations A slow website requires 50% concentration of additional efforts 78 % of users feel stressed or angry when using a slow website Increase in usage time for employees Sources :Jakob Nielsen : http://www.useit.com/alertbox/response-times.html, Foviance: http://www.foviance.com/what-we-think/wen- stress-a-wake-up-call-for european-business/, Harris interactive: http://www.tealeaf.com/news/news-releases/2011/tealeaf-announces- new-mobile-research.php, PhoCusWright : http://www.phocuswright.com/free_reports/consumer-response-to-travel-site-performance, Redhat http://www.webperformancetoday.com/2010/07/01/the-best-graphs-of-velocity/
  • 7. 7 Why is performance important ? The Impact of mobility How to test ? Yesterdays solutions Todays Solutions
  • 8. 8 The sources of performance Page construction Answers Request Display Treatment Performance = Device processing + Time to communication + Server processing Latency Debit Amount of Data Quantity of requests Protocol
  • 9. 9 TCP SYN « Hi !» TCP SYN ACK « Hi ?! » TCP ACK « There you are ! » HTTP GET « I want this » HTTP 200« OK » [index.html] « Here it is : ….» Note: Secured version (SSL ), the dialogue before data recovery requires a turn and return to give The " encryption keys " HTTP The most common protocol (HTTP) = 2 round trips Time for one round trip = latency
  • 10. Estimated latency of the different types of network 10Source : http://www.digitalsociety.org/2010/08/conflating-broadband-speed-with-internet-speed-is-misleading/ http://www.dslreports.com/ Because 2 round trips are required by HTTP, first « useful » data reaches application in 40ms for ADSL , in 300ms in 3G 0 100 200 300 400 EDGE/2G 3G Optical fiber inter-continental Optical fiber inter-country ADSL Private local Network
  • 11. 11 4G delivers us ? 11http://opensignal.com/reports/state-of-lte-q1-2014/ Yes, it multiple the download by 4 More dedicated bandwidth depends on the overall traffic and mobile traffic increased by 68 % in France But it reduces latency just by 40 %
  • 12. 12 Why is performance important ? The Impact of mobility How to test ? Yesterdays solutions ? Todays solutions ?
  • 13. 13 Tools to analyse application performance (APM : Application Performance Management) Synthetic testing Real User Monitoring (RUM) Probe to integrate into applications or based on emulations Objectives : crash, display time , ... by screens / by request Actors: Compuware Riverbet , AppDynamics , New Relic , Akamai ( Movbitest ) , Keynote Systems, Witbe, Dynatrace, Aternity, Newrelic, … Robots testing applications on devices
  • 14. 14 Why is performance important ? The impact of mobility How to test ? Yesterdays solution Todays solutions
  • 15. 15 Time to communication is a function of Perception, latency, debit, Amount of data, Number of requests, Protocol.
  • 16. 16 Display items quickly Programmation : flush , Post-load , pre-load CDN Akamai Technologies, Edgecast, level 3, Amazon, OVH, CDNetworks Cache memories At the server level and devices Prepare ahead cache Infrastructures Bandwidth & CPU Compress Gzip/deflate via accept-encoding / content-encoding Data formats Ex : JSON > HTML Binary communication protocol To reduce the amount of data to be transmitted Programming Quality Reduce the number of request by grouping , Putting scripts after the maximum requests , Avoid interdependent queries, Web: avoid CSS expressions, Redirects , ... . Time to communication is a function of Perception, latency, debit, Amount of data, Number of requests, Protocol.
  • 17. 17 Why is performance important ? The impact of mobility How to test ? Yesterdays solutions Todays solutions
  • 18. 18 Latency 10 to 20ms Latency 80 to 160 ms Source of the mobility problem
  • 19. 19 19 1st Goal: Reduce the amount of data to be transferred to the mobile 2nd goal: To reduce the number of round trips for the same amount of data Source of the mobility problem Latency 10 to 20ms Latency 80 to 160 ms
  • 20. 20 Sources Server « middleware mobile » Agregation Cache Cache Orchestration Eliminate unnecessary data Avoid the multiple calls from the mobile (*) Send only new data Using bidirectional protocol (*) Especially the application requires interdependent calls ( if the value of A to be found in such a server is X, then search B in another server)
  • 21. 21 « Hello !» « Yes ?! » « Cool ! » « I want this » « OK » « Here! » HTTP Page 1 Page 2 « Hello !» « Again?! » « Cool ! » « I want this » « OK » « Here! » Page 1 – Data refresh (*) « Yeah … » « Cool !» « OK » « Here! » « Hello !» « Yes ?! » « I want this » « OK » « Tiens! » Page 1 2nd page «I want this » « Here! » 1st page – data refresh « Here! » Bidirectional protocol Note: (* ) upgradeable scenario with long polling Compared to the conventional protocol and two-way protocol « Hello !» « I want this » « Cool ? »
  • 22. 22 Transferred data for the same transaction Mobile application for trading Transferred data for the same transaction Mobile application for sports betting 298 95 1.6 0 50 100 150 200 250 300 350 Web app Mobile app Motwin app Ko Ko 608 243 48 0 100 200 300 400 500 600 700 Web app Mobile app Motwin app -99% -92% The impact of these new solutions on the quantity of transferred data
  • 23. 23 Do It Yourself or BUY/RENT Motwin Platform
  • 24. 24 Implementation of the Motwin platform
  • 25. 25 3 sensitive segments performance Trading & Betting • Access to real-time data • UI faster for customers in a hurry to buy / sell / bet Collaborators Tools • Real time CRM/ERP • Assisted maintenance IoT real time BtB • Configuration, Maintenance and Monitoring • IU efficient and real - time control objects 25 Critical data Efficiency and Productivity Critical systems
  • 26. 26 Reactivity Disponibility Speed deployment Maintenance Cost and Scalability -30 to 70% 90% less data Management Microcuts x 7 Beyond the first OS -30% to -60%
  • 27. Thank You ! Stand : E32B 27 community@motwin.com www.motwin.com @motwin_inc