William Louth    JINSPIRED      @jxinsight www.jinspired.comwilliam@jinspired.com
Data
LocalRemote   Past   Present   Predicted
Information   Management   Model        Model
Measurement
Method      Metering      Metric Analysis    Causality   Correlation Accuracy     Event       SampledAllocation    Thread ...
Model     Activity ResourceDevice     Probe      MeterDevelop    Code      CounterDesign    Behavior    Usage Data      Gr...
Complexity
Space                        Complexity               ity             rs          ve                  ism        Di       ...
HTTP                                                                                      API                             ...
Client    sf://.......cl:clock.time=                   cl   Response                        sf:clock.time=                ...
Client  sf://..........cl:clock.time=                    cl                         ☁ Response                         sf:...
ControlAwareness& Adaption
Profile   Protect   Police   Prioritize   Predict   Provision
Self Adaptive SoftwareSelf Adaptive Software evaluates its own behaviorand changes behavior when the evaluationindicates t...
Evidence                                                        Observation                  1                            ...
DisturbancesGoals        Control            Process                  Sensor
public int func(...) {                                      QoS                    reserve(func)   Resource..................
User   WebPage   Resource
SystemDynamics
+                  _Inflow                Stock               Outflow + _                                        + _      ...
+                     _              +                     _Thread Pool              Reserved        Concurrency          ...
operations            System      { system execution model }   Dynamics                                                   ...
Finally....Faster
Sweeper                                                                       (Secondary Processor)                       ...
Jinspired june2012
Upcoming SlideShare
Loading in...5
×

Jinspired june2012

399

Published on

This presentations shows how applications & services are measured, monitored even controlled in the cloud along with some novel approaches.

Published in: Technology, Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
399
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Jinspired june2012

  1. 1. William Louth JINSPIRED @jxinsight www.jinspired.comwilliam@jinspired.com
  2. 2. Data
  3. 3. LocalRemote Past Present Predicted
  4. 4. Information Management Model Model
  5. 5. Measurement
  6. 6. Method Metering Metric Analysis Causality Correlation Accuracy Event SampledAllocation Thread ProcessAssignment Direct Apportioned
  7. 7. Model Activity ResourceDevice Probe MeterDevelop Code CounterDesign Behavior Usage Data Group Metering
  8. 8. Complexity
  9. 9. Space Complexity ity rs ve ism Di m na Dy Time
  10. 10. HTTP API ta da Cloud data/code Service dataApplication HTTP Application Activity API data da ta Cloud Cloud code Service Service/Shell Activity Cloud Service/Shell
  11. 11. Client sf://.......cl:clock.time= cl Response sf:clock.time= sf:cpu.time= sf:io.bytes= sf:charge.unit= ☁ sf ☁ ☁ salesforce.comsf:clock.time=sf:cpu.time=sf:io.bytes=sf:charge.unit= Metering Management Service
  12. 12. Client sf://..........cl:clock.time= cl ☁ Response sf:clock.time= sf:cpu.time= sf:io.bytes= sf:charge.unit= sf:db.time= sf:db.count= s3:clock.time= s3:io.bytes= s3:charge.unit= ☁ ☁ sf salesforce.com db Response ☁sf:clock.time= s3:clock.time= ☁sf:cpu.time= s3:io.bytes=sf:io.bytes= ☁ s3:charge.unit=sf:charge.unit= s3sf:db.time=sf:db.count=s3:clock.time= aws.amazon.coms3:io.bytes=s3:charge.unit= MeteringManagement Service
  13. 13. ControlAwareness& Adaption
  14. 14. Profile Protect Police Prioritize Predict Provision
  15. 15. Self Adaptive SoftwareSelf Adaptive Software evaluates its own behaviorand changes behavior when the evaluationindicates that it is not accomplishing what thesoftware is intended to do, or when betterfunctionality or performance is possible.” DARPA
  16. 16. Evidence Observation 1 Self RegulatedAction 4 Feedback Loop 2 Relevance Reaction Judgement 3 Consequence
  17. 17. DisturbancesGoals Control Process Sensor
  18. 18. public int func(...) { QoS reserve(func) Resource................................ Rate...... Limiting........... Priority..... Queueing... release(units)} Reservation Lanes
  19. 19. User WebPage Resource
  20. 20. SystemDynamics
  21. 21. + _Inflow Stock Outflow + _ + _ Sensor + _ Births Population Deaths + + reinforcing balancing loop loop
  22. 22. + _ + _Thread Pool Reserved Concurrency Released initial size + _ + _QoS Resource Reserved QoS Reservation Released initial capacity
  23. 23. operations System { system execution model } Dynamics Application unification System Dynamics Software development Application{ software execution model } Software
  24. 24. Finally....Faster
  25. 25. Sweeper (Secondary Processor) Reduce Friction Increase the Speed Thrower Trajectory Curling Stone Target[Primary Processor] Predicted Path Reduce Curl Straighten the Path Shorten Sweeper (Secondary Processor)

×