1 / 26

Florin Dinu T. S. Eugene Ng Rice University

Synergy2Cloud: Introducing Cross- Sharing of Application Experiences Into the Cloud Management Cycle. Florin Dinu T. S. Eugene Ng Rice University. Multi-Tenant Clouds . Resource contention Performance variation Failures. App1. App2.

arden
Download Presentation

Florin Dinu T. S. Eugene Ng Rice University

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Synergy2Cloud: Introducing Cross-Sharing of Application Experiences Into the Cloud Management Cycle Florin Dinu T. S. Eugene Ng Rice University

  2. Multi-Tenant Clouds • Resource contention • Performance variation • Failures App1 App2 Challenging to ensure good application performance

  3. Our Position: Cross-Sharing Experiences ? ! Red would benefit ifGreenshared info about the failure

  4. Ensuring Performance with Good Decisions Good provisioning decisions App1: How to best scale next? Need information to guide provisioning decisions

  5. Ensuring Performance with Good Decisions Good runtime decisions Detect node failure & restart tasks Need information to guide runtimedecisions

  6. Ensuring Performance with Good Decisions • Generating an efficient execution plan [NSDI 12] • Scheduling a known execution plan [SOSP 09] • Runtime decisions • Stragglers [OSDI 08, OSDI 10] • Managing traffic [SIGCOMM 2011] • Failures [HPDC 12] All these decisions can benefit from extra information

  7. Obtaining Information Today: Operator App 1 App 2 Careful on resources Application agnostic data Operator monitoring can be useful but has important shortcomings

  8. Obtaining Information Today: Apps Limited view of environment Limited to owned resources Limited by scale of app Application monitoring also has several limitations

  9. Commonality among cloud apps LARGE SCALE FWK e.g. MapReduce. Many apps built on top OS Few OS image types available HARDWARE Virtualized. Few instance types available Unique opportunity in the cloud

  10. Cross-Sharing Application Experiences Sharing I need to scale out fast Benefiting from information shared by others

  11. Today: Rudimentary Sharing Too slow. Requires human involvement. Needs to be automatic.

  12. Hurdle to Overcome: Isolation Isolation impedes sharing and measurement

  13. But Isolation is Important • More predictable performance • Improved security • Today, we are striving to isolate at all levels • Network [NSDI 11, CONEXT 10, Xen] • Storage [VMWare] • CPU [Xen] • Caches [CCSW 09] Is complete isolation the way to go?

  14. Cross-Sharing Application Experiences Examples Incentives Challenges

  15. Cross-Sharing Example: Performance Dedicated Storage Different cloud Latency Throughput Shared performance information may helpscheduling

  16. Cross-Sharing Example: Scalability • Red needs to scale fast • No time for testing • Which VM type to use? • Information from green • may be useful ? Sharing can inform scaling decisions

  17. Cross-Sharing Example: Failures ? • Detect compute node failures • Green deployed at larger scale • Green detects failure faster • Red has few vantage points • Red can benefit from • green’s experience ! Some applications are better suited to discover failures

  18. Incentives for the Operator • 1. App performance = cloud performance • 2. Apps = huge monitoring platform • 3. $$$ • 4. If you can’t beat them, join them Share 3rd party Use

  19. Incentives for the Applications • 1. More info => better decisions • 2. Obtain info quickly – no trial and error needed • 3. Info about resources that they don’t use • 4. Some apps better suited than others ? !

  20. Addressing Challenges • Increasing information expressiveness • Verifying information authenticity • Establishing similarity between applications Synergy between applications and operator

  21. Challenge – Increasing Expressiveness Path is congested Another Data Center Limited infrastructure knowledge

  22. Challenge – Increasing Expressiveness What does 100% CPU mean for me? Empty slots 100% CPU usage Effects of isolation – Info meaningul in context of red only Operator should provide abstract location information

  23. Challenges - Authenticity Storage I have 20 instances connecting simultaneously to storage and my throughput is…. • Verify where possible (ownership) • Cull outliers with statistics • Filter incoming sources of shared data Fake Performance Info vs Bad Performance Info

  24. Challenges – Establishing Similarity • Difficult in the general case • Not all layers impact every • shareable information (compute failures) • Detect paths in the layers that impactthe shared information • Encode similarity between paths (e.g. hash codes) Small changes in the code LARGE SCALE FWK Configuration, code tweaks OS Use delayed ACK? HARDWARE Impact runtime ALLOCATION 50 vs 3 VMs

  25. Synergy2Cloud.com Our first step for making cross-sharing a reality

  26. Conclusion • Our position: cross-sharingexperiences a promising direction • Win-win for both applications and operator • Not trivial - several interesting challenges ahead

More Related