SlideShare une entreprise Scribd logo
1  sur  51
Télécharger pour lire hors ligne
Object graphs
swapping
m a r i a n o m a r t i n e z p e c k
m a r i a n o p e c k @ g m a i l . c o m
RMod
1
Wednesday, September 15, 2010
The context
2
Wednesday, September 15, 2010
The context
2
Wednesday, September 15, 2010
The context
2
Wednesday, September 15, 2010
The context
2
Wednesday, September 15, 2010
The context
2
Wednesday, September 15, 2010
The context
2
Wednesday, September 15, 2010
Problem
Use more memory than needed.
Make OOP languages unsuitable for memory
limited devices.
Existence of unused but referenced objects.
3
Wednesday, September 15, 2010
The context
In OOP primary memory is represented by an object graph
A B C
D E F G H
Y
K
X
I LJ
Z
4
Wednesday, September 15, 2010
Garbage Collector
Only collects objects that nobody else points to.
A B C
D E F G H
Y
K
X
I LJ
Z
5
Wednesday, September 15, 2010
Garbage Collector
Only collects objects that nobody else points to.
A B C
D E F G H
Y
K
X
I LJ
Z
5
Wednesday, September 15, 2010
Garbage Collector
Only collects objects that nobody else points to.
A B C
D E F G H
Y
K
X
I LJ
Z
5
Wednesday, September 15, 2010
Garbage Collector
Only collects objects that nobody else points to.
A B C
D E F G H
Y
K
X
I LJ
Z
5
Wednesday, September 15, 2010
6
But...what happens with referenced yet unused objects?
A B C
D E F G H
Y
K
X
I LJ
Z
Wednesday, September 15, 2010
Idea
Swap out (not remove) unused objects to
disk.
Automatically load them back when needed.
7
Wednesday, September 15, 2010
Related work
Large object oriented memory (LOOM).
Melt - Supporting memory leaks.
ImageSegments.
8
Wednesday, September 15, 2010
But...no one solves all the problems
9
Wednesday, September 15, 2010
Main challenges
Not to use more memory than the one
released by swapping.
Low overhead penalty.
Group objects in an smart way.
10
Wednesday, September 15, 2010
Key aspects
Mark and trace unused/used objects at
runtime.
The usage of proxies.
Group unused objects (subgraphs).
11
Wednesday, September 15, 2010
Why we need to
group objects?
Because if we replace each object by a proxy,
we release little memory.
We want to replace a whole group by one or a
few proxies.
12
Wednesday, September 15, 2010
Why subgraphs?
Group of objects that are used (or not used)
together.
We need few proxies (for the roots) for
several objects.
13
Wednesday, September 15, 2010
Experiments done
Modify Smalltalk VM to mark and trace
objects usage.
Visualize objects and memory usage.
Take statistics from different scenarios.
14
Wednesday, September 15, 2010
deployed web
application example
4%
96%
Amount of objects
Used Unused
15%
85%
Amount of memory
Used Unused
15
Wednesday, September 15, 2010
Swapping steps and
challenges
1.Identify sets of objects and serialize them.
Problems: cycles, speed, etc.
2.Write the serialized objects into a file.
Problems: file format, encoding, speed, etc.
3. Load the objects from a file. Problems: class
reshape, avoid duplicates, speed, etc.
16
Wednesday, September 15, 2010
Subgraphs
A
(Root)
B
(Root)
C
(Root)
D E F G H
Y
K
X
I LJ
Subgraph to process
Z
Roots Inner Shared External
17
Wednesday, September 15, 2010
More
problems
Should shared objects be
included or not?
GC moves objects.
Pointers update.
Class changes.
Recreate and reinitialize
objects.
Code executed after
loading.
A
(Root)
B
(Root)
C
(Root)
D E F G H
Y
K
X
I LJ
Subgraph to process
Z
Roots Inner Shared External
18
Wednesday, September 15, 2010
ImageSegment
Wednesday, September 15, 2010
ImageSegment basis
Only write/swap roots and inner objects.
Shared objects are NOT swapped.
Keep an array in memory for the shared
objects.
Update object pointers to point to a relative
address inside the arrays (offset).
Roots are replaced by proxies.
Uses GC facilities to detect shared objects.
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
D
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
D
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’
D
offset
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’ E’
D
offset
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’ E’
D
offset
offset
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’ E’
D
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’ E’
D
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
Subgraph traverse
A CB
D E F G H
I J K L
A’ B’ C’ E’
F GD J
I’ H’ L’
Serialized objects WordArray
Shared objects Array
Wednesday, September 15, 2010
A CB
D E F G H
I J K L
A’ B’ C’ E’
F GD J
I’ H’ L’
Serialized objects WordArray
Shared objects Array
Offset Memory address
Wednesday, September 15, 2010
A CB
D E F G H
I J K L
A’ B’ C’ E’ F GD JI’ H’ L’
Serialized objects WordArray Shared objects Array
anImageSegment
Wednesday, September 15, 2010
D E F G H
I J K L
A’ B’ C’ E’ F GD JI’ H’ L’
Serialized objects WordArray Shared objects Array
anImageSegment
P1 P3P2
Wednesday, September 15, 2010
D E F G H
I J K L
A’ B’ C’ E’ F GD JI’ H’ L’
Serialized objects WordArray Shared objects Array
anImageSegment
P1 P3P2
Wednesday, September 15, 2010
D F G
J K
A’ B’ C’ E’ F GD JI’ H’ L’
Serialized objects WordArray Shared objects Array
anImageSegment
P1 P3P2
Wednesday, September 15, 2010
CB
D F G
J K
A’ B’ C’ E’ I’ H’ L’
Serialized objects WordArray
F GD J
Shared objects Array
anImageSegment
P1 P3P2
Binary file
Wednesday, September 15, 2010
CB
D F G
J K
F GD J
Shared objects Array
anImageSegment
P1 P3P2
Binary file
A’ B’ C’ E’ I’ H’ L’
Wednesday, September 15, 2010
CB
D F G
J K
F GD J
Shared objects Array
anImageSegment
P1 P3P2
Binary file
A’ B’ C’ E’ I’ H’ L’
NIL
Wednesday, September 15, 2010
ImageSegment
conclusions
Good speed.
Graph traverse is done in VM side.
Good use of GC facilities.
You have to be aware of shared objects.
Bad granularity level.
Implicit needed information in object graphs.
Wednesday, September 15, 2010
Thanks!
Mariano Martinez Peck
marianopeck@gmail.com
Wednesday, September 15, 2010

Contenu connexe

Similaire à Object garphs swapping

Seaside - Why should you care? (Dynamic Stockholm 2010)
Seaside - Why should you care? (Dynamic Stockholm 2010)Seaside - Why should you care? (Dynamic Stockholm 2010)
Seaside - Why should you care? (Dynamic Stockholm 2010)jfitzell
 
RIA Unleashed - Developing for the TV with litl os
RIA Unleashed - Developing for the TV with litl osRIA Unleashed - Developing for the TV with litl os
RIA Unleashed - Developing for the TV with litl osryancanulla
 
JVM Internals - NEJUG Nov 2010
JVM Internals - NEJUG Nov 2010JVM Internals - NEJUG Nov 2010
JVM Internals - NEJUG Nov 2010Doug Hawkins
 
Poisoning Rubinius: The _why and How
Poisoning Rubinius: The _why and HowPoisoning Rubinius: The _why and How
Poisoning Rubinius: The _why and HowBrian Ford
 
Mastering CSS3 Selectors
Mastering CSS3 SelectorsMastering CSS3 Selectors
Mastering CSS3 SelectorsRachel Andrew
 
Modern web application model
Modern web application modelModern web application model
Modern web application modelMichal Taberski
 
Impactos no design com programação funcional
Impactos no design com programação funcionalImpactos no design com programação funcional
Impactos no design com programação funcionalLuiz Costa
 

Similaire à Object garphs swapping (8)

Seaside - Why should you care? (Dynamic Stockholm 2010)
Seaside - Why should you care? (Dynamic Stockholm 2010)Seaside - Why should you care? (Dynamic Stockholm 2010)
Seaside - Why should you care? (Dynamic Stockholm 2010)
 
RIA Unleashed - Developing for the TV with litl os
RIA Unleashed - Developing for the TV with litl osRIA Unleashed - Developing for the TV with litl os
RIA Unleashed - Developing for the TV with litl os
 
JVM Internals - NEJUG Nov 2010
JVM Internals - NEJUG Nov 2010JVM Internals - NEJUG Nov 2010
JVM Internals - NEJUG Nov 2010
 
Poisoning Rubinius: The _why and How
Poisoning Rubinius: The _why and HowPoisoning Rubinius: The _why and How
Poisoning Rubinius: The _why and How
 
06 data
06 data06 data
06 data
 
Mastering CSS3 Selectors
Mastering CSS3 SelectorsMastering CSS3 Selectors
Mastering CSS3 Selectors
 
Modern web application model
Modern web application modelModern web application model
Modern web application model
 
Impactos no design com programação funcional
Impactos no design com programação funcionalImpactos no design com programação funcional
Impactos no design com programação funcional
 

Plus de Mariano Martínez Peck

Plus de Mariano Martínez Peck (6)

Clustered Serialization with Fuel
Clustered Serialization with FuelClustered Serialization with Fuel
Clustered Serialization with Fuel
 
PhD Thesis Defense Presentation
PhD Thesis Defense PresentationPhD Thesis Defense Presentation
PhD Thesis Defense Presentation
 
Building you own Pharo images with Metacello
Building you own Pharo images with Metacello Building you own Pharo images with Metacello
Building you own Pharo images with Metacello
 
Visualizing Objects and Memory Usage
Visualizing Objects and Memory UsageVisualizing Objects and Memory Usage
Visualizing Objects and Memory Usage
 
Object swapping issues and the imagesegment implementation
Object swapping issues and the imagesegment implementationObject swapping issues and the imagesegment implementation
Object swapping issues and the imagesegment implementation
 
SqueakDBX
SqueakDBXSqueakDBX
SqueakDBX
 

Dernier

Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Igalia
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityPrincipled Technologies
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Scriptwesley chun
 
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEarley Information Science
 
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...Neo4j
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking MenDelhi Call girls
 
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...gurkirankumar98700
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slidevu2urc
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilV3cube
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...Martijn de Jong
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slidespraypatel2
 
A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024Results
 
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxKatpro Technologies
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘RTylerCroy
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 

Dernier (20)

Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
 
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men
 
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of Brazil
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slides
 
A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024
 
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 

Object garphs swapping

  • 1. Object graphs swapping m a r i a n o m a r t i n e z p e c k m a r i a n o p e c k @ g m a i l . c o m RMod 1 Wednesday, September 15, 2010
  • 8. Problem Use more memory than needed. Make OOP languages unsuitable for memory limited devices. Existence of unused but referenced objects. 3 Wednesday, September 15, 2010
  • 9. The context In OOP primary memory is represented by an object graph A B C D E F G H Y K X I LJ Z 4 Wednesday, September 15, 2010
  • 10. Garbage Collector Only collects objects that nobody else points to. A B C D E F G H Y K X I LJ Z 5 Wednesday, September 15, 2010
  • 11. Garbage Collector Only collects objects that nobody else points to. A B C D E F G H Y K X I LJ Z 5 Wednesday, September 15, 2010
  • 12. Garbage Collector Only collects objects that nobody else points to. A B C D E F G H Y K X I LJ Z 5 Wednesday, September 15, 2010
  • 13. Garbage Collector Only collects objects that nobody else points to. A B C D E F G H Y K X I LJ Z 5 Wednesday, September 15, 2010
  • 14. 6 But...what happens with referenced yet unused objects? A B C D E F G H Y K X I LJ Z Wednesday, September 15, 2010
  • 15. Idea Swap out (not remove) unused objects to disk. Automatically load them back when needed. 7 Wednesday, September 15, 2010
  • 16. Related work Large object oriented memory (LOOM). Melt - Supporting memory leaks. ImageSegments. 8 Wednesday, September 15, 2010
  • 17. But...no one solves all the problems 9 Wednesday, September 15, 2010
  • 18. Main challenges Not to use more memory than the one released by swapping. Low overhead penalty. Group objects in an smart way. 10 Wednesday, September 15, 2010
  • 19. Key aspects Mark and trace unused/used objects at runtime. The usage of proxies. Group unused objects (subgraphs). 11 Wednesday, September 15, 2010
  • 20. Why we need to group objects? Because if we replace each object by a proxy, we release little memory. We want to replace a whole group by one or a few proxies. 12 Wednesday, September 15, 2010
  • 21. Why subgraphs? Group of objects that are used (or not used) together. We need few proxies (for the roots) for several objects. 13 Wednesday, September 15, 2010
  • 22. Experiments done Modify Smalltalk VM to mark and trace objects usage. Visualize objects and memory usage. Take statistics from different scenarios. 14 Wednesday, September 15, 2010
  • 23. deployed web application example 4% 96% Amount of objects Used Unused 15% 85% Amount of memory Used Unused 15 Wednesday, September 15, 2010
  • 24. Swapping steps and challenges 1.Identify sets of objects and serialize them. Problems: cycles, speed, etc. 2.Write the serialized objects into a file. Problems: file format, encoding, speed, etc. 3. Load the objects from a file. Problems: class reshape, avoid duplicates, speed, etc. 16 Wednesday, September 15, 2010
  • 25. Subgraphs A (Root) B (Root) C (Root) D E F G H Y K X I LJ Subgraph to process Z Roots Inner Shared External 17 Wednesday, September 15, 2010
  • 26. More problems Should shared objects be included or not? GC moves objects. Pointers update. Class changes. Recreate and reinitialize objects. Code executed after loading. A (Root) B (Root) C (Root) D E F G H Y K X I LJ Subgraph to process Z Roots Inner Shared External 18 Wednesday, September 15, 2010
  • 28. ImageSegment basis Only write/swap roots and inner objects. Shared objects are NOT swapped. Keep an array in memory for the shared objects. Update object pointers to point to a relative address inside the arrays (offset). Roots are replaced by proxies. Uses GC facilities to detect shared objects. Wednesday, September 15, 2010
  • 29. Subgraph traverse A CB D E F G H I J K L Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 30. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 31. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 32. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 33. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 34. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ D Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 35. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ D Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 36. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ D offset Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 37. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ E’ D offset Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 38. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ E’ D offset offset Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 39. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ E’ D Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 40. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ E’ D Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 41. Subgraph traverse A CB D E F G H I J K L A’ B’ C’ E’ F GD J I’ H’ L’ Serialized objects WordArray Shared objects Array Wednesday, September 15, 2010
  • 42. A CB D E F G H I J K L A’ B’ C’ E’ F GD J I’ H’ L’ Serialized objects WordArray Shared objects Array Offset Memory address Wednesday, September 15, 2010
  • 43. A CB D E F G H I J K L A’ B’ C’ E’ F GD JI’ H’ L’ Serialized objects WordArray Shared objects Array anImageSegment Wednesday, September 15, 2010
  • 44. D E F G H I J K L A’ B’ C’ E’ F GD JI’ H’ L’ Serialized objects WordArray Shared objects Array anImageSegment P1 P3P2 Wednesday, September 15, 2010
  • 45. D E F G H I J K L A’ B’ C’ E’ F GD JI’ H’ L’ Serialized objects WordArray Shared objects Array anImageSegment P1 P3P2 Wednesday, September 15, 2010
  • 46. D F G J K A’ B’ C’ E’ F GD JI’ H’ L’ Serialized objects WordArray Shared objects Array anImageSegment P1 P3P2 Wednesday, September 15, 2010
  • 47. CB D F G J K A’ B’ C’ E’ I’ H’ L’ Serialized objects WordArray F GD J Shared objects Array anImageSegment P1 P3P2 Binary file Wednesday, September 15, 2010
  • 48. CB D F G J K F GD J Shared objects Array anImageSegment P1 P3P2 Binary file A’ B’ C’ E’ I’ H’ L’ Wednesday, September 15, 2010
  • 49. CB D F G J K F GD J Shared objects Array anImageSegment P1 P3P2 Binary file A’ B’ C’ E’ I’ H’ L’ NIL Wednesday, September 15, 2010
  • 50. ImageSegment conclusions Good speed. Graph traverse is done in VM side. Good use of GC facilities. You have to be aware of shared objects. Bad granularity level. Implicit needed information in object graphs. Wednesday, September 15, 2010