an AI approach toweb gaming in the cloud   Vittorio Amos Ziparo        @OWF 2011
Cloud-TMco-financed by the European Commission through the contract no. 257784.
Web GamesBrowser Games      Flash Games  Gameforge > 300M reg users!Zynga has 50% of facebook users!
Outlook●   an AI design pattern●   scaling in the Cloud
an AIdesign pattern
MAS: an AI pattern             Percept                         more agents             ActionAgent      Environment
Why MAS?1. Benefits of a cmd pattern2. Guarantees on correctness3. Easy to plug in AI and MAS code
MadMass            rc  e       so  u          e r )  p en           e r vo            ( s            )      y lib         ...
scaling inthe Cloud
The price of scaling
A DSTM data layerDistributed Software Transactional Memory     ... imagine a memory that elastically scales,          that...
A Red Hat Backbone
Autonomic Self-Optimization
Want to know more?            oludap.comgithub.com/algorithmica/madmass           cloudtm.euwww.hibernate.org/subprojects/...
Thanks!
Cloud / Oludap case study - Vittorio Amos Ziparo, Algorithmica
Cloud / Oludap case study - Vittorio Amos Ziparo, Algorithmica
Upcoming SlideShare
Loading in...5
×

Cloud / Oludap case study - Vittorio Amos Ziparo, Algorithmica

653

Published on

Oludap is a multiplayer online strategic game that is based on the open-source (j)Ruby on Rails framework MADMASS (MAssively Distributed Multi Agent System Simulator). MADMASS simplifies the development of next-generation web applications by relying on methodologies from Artificial Intelligence and Multi-Agent Systems.

In Oludap, players compete in real-time for limited resources in a shared environment. Oludap’s gameplay is episodic, meaning that small groups of users can play one-shot games. This type of games presents two challenges:

the episodic nature of the game-play may yield to rapid and unpredictable fluctuations in the workload that can make planning for IT infrastructures not sustainable;
the fact that players act on a shared environment generates a high level of contention on data thus, considering that consistency is a strong requirement, scalability may be at risk.

In order to overcome these challenges, we are working to re-implement the data layer of the game on top of Hibernate OGM and Infinispan, in the framework of the Specific Targeted Research Project (STReP) Cloud-TM, which is co-financed by the European Commission through the contract no. 257784.

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
653
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
2
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Cloud / Oludap case study - Vittorio Amos Ziparo, Algorithmica

  1. 1. an AI approach toweb gaming in the cloud Vittorio Amos Ziparo @OWF 2011
  2. 2. Cloud-TMco-financed by the European Commission through the contract no. 257784.
  3. 3. Web GamesBrowser Games Flash Games Gameforge > 300M reg users!Zynga has 50% of facebook users!
  4. 4. Outlook● an AI design pattern● scaling in the Cloud
  5. 5. an AIdesign pattern
  6. 6. MAS: an AI pattern Percept more agents ActionAgent Environment
  7. 7. Why MAS?1. Benefits of a cmd pattern2. Guarantees on correctness3. Easy to plug in AI and MAS code
  8. 8. MadMass rc e so u e r ) p en e r vo ( s ) y lib e n t r u b ( c li li b js
  9. 9. scaling inthe Cloud
  10. 10. The price of scaling
  11. 11. A DSTM data layerDistributed Software Transactional Memory ... imagine a memory that elastically scales, that transparently persists data, and that deals for you with concurrency issues
  12. 12. A Red Hat Backbone
  13. 13. Autonomic Self-Optimization
  14. 14. Want to know more? oludap.comgithub.com/algorithmica/madmass cloudtm.euwww.hibernate.org/subprojects/ogm www.jboss.org/infinispan torquebox.org
  15. 15. Thanks!
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×