Dicas e truques para ser um bom inquilino no Cloud
4 likes1,353 views
1) The document provides tips and tricks for being a good cloud tenant, discussing concepts like spreading processes across containers and zones for better performance and utilization of resources. 2) It discusses log routing and processing techniques like using logplex to collect logs and fluentd to handle event processing and metric collection. 3) The document recommends techniques for log storage, analysis and monitoring including using services like Librato, Heroku Pager and external monitoring tools to get insights from logs, metrics and events.
![weighted
fair scheduling
[1]](https://image.slidesharecdn.com/platform-multitenancy-qconsp-120804165252-phpapp01/85/Dicas-e-truques-para-ser-um-bom-inquilino-no-Cloud-7-320.jpg)
![CPU stacking [2]](https://image.slidesharecdn.com/platform-multitenancy-qconsp-120804165252-phpapp01/85/Dicas-e-truques-para-ser-um-bom-inquilino-no-Cloud-22-320.jpg)
![I/O intensive
applications need their
own domain [3]](https://image.slidesharecdn.com/platform-multitenancy-qconsp-120804165252-phpapp01/85/Dicas-e-truques-para-ser-um-bom-inquilino-no-Cloud-24-320.jpg)
Dicas e truques para ser um bom inquilino no Cloud
- 1. dicas e truques para ser um bom inquilino ... no Cloud Fabio Kung
- 3. inquilino?
- 4. A B C D multitenancy
- 5. IaaS PaaS Containers/Zones
- 6. considere ... espalhar seus processos
- 7. weighted fair scheduling [1]
- 8. App A P1 P2 P3 20 App B P4 P5 P6 P7 40 App C P8 40
- 9. App A App B App C 1 2 3 0 4 5 6 7 0 8 0 20 40 40 fator de “justiça”
- 10. App B App C 4 5 6 7 0 8 0 40 40
- 11. App B App C App A 4 5 6 7 0 8 0 1 2 3 30 40 40 20
- 12. App C App A 8 0 1 2 3 30 40 20
- 13. App C App B App A 8 0 4 5 6 7 40 1 2 3 30 40 40 20
- 14. App B App A 4 5 6 7 40 1 2 3 30 40 20
- 15. App B App A App C 4 5 6 7 40 1 2 3 30 8 80 40 20 40
- 16. App A App C 1 2 3 30 8 80 20 40 minimum + bursting
- 17. co-scheduling
- 18. SMP
- 19. VM A VM B VM C vCPU vCPU vCPU vCPU vCPU
- 20. synchronization costs
- 21. VM A VM B VM C vCPU vCPU vCPU vCPU vCPU
- 22. CPU stacking [2]
- 23. evite I/O ... se puder
- 24. I/O intensive applications need their own domain [3]
- 25. espalhe seus processos http://www.12factor.net/concurrency
- 26. efemeridade
- 27. eu acho...
- 28. PROVE!
- 29. blá blá blá ... monitoração ... blá blá blá
- 30. logs => event stream
- 31. pipeline Archive, dashboards inc App id logs en ts t s even Log App metrics logs router Pager eve nt s logs App Event nts ide processor inc
- 32. Log routers • https://github.com/heroku/logplex • https://github.com/fluent/fluentd events-to-metrics • https://github.com/heroku/pulse
- 37. Librato
- 38. the pager
- 41. pipeline Archive, dashboards inc App id logs en ts t s even Log App metrics logs router Pager eve nt s logs App Event nts ide processor inc
- 42. pipeline Archive, dashboards inc App id en ts t s even External App metrics events Monitor Pager eve nt s App Event nts ide processor in c
- 44. média
- 46. %til
- 49. naked domains https://devcenter.heroku.com/articles/avoiding-naked-domains-dns-arecords
- 51. References 1. Groves, Taylor, Jeff Knockel and Eric Schulte. “BFS vs. CFS - Scheduler Comparison”. 11 Dec 2009. 2 Aug 2012 <http:// slimjim.cs.unm.edu/~eschulte/data/bfs-v-cfs_groves-knockel- schulte.pdf> 2. Sukwong, Orathai and Hyong S. Kim. “Is Co-scheduling Too Expensive for SMP VMs?”. EuroSys 2011 12 Apr 2011, Carnegie Mellon University. 29 Jul 2012 <http://eurosys2011.cs.uni-salzburg.at/pdf/ eurosys2011-sukwong-slides.pdf> 3. Ongaro, Diego, Alan L. Cox and Scott Rixner. “Scheduling I/O in Virtual Machine Monitors”. Rice University. 2 Aug 2012 <http:// www.cs.rice.edu/CS/Architecture/docs/ongaro-vee08.pdf> 4. Pijewski, Bill. “Our ZFS I/O Throttle”. Bill Pijewski's Blog. March 1, 2011. August 3, 2012 <http://dtrace.org/blogs/wdp/2011/03/ our-zfs-io-throttle/>
Editor's Notes
- #2: \n
- #3: \n
- #4: \n
- #5: unidades computacionais\n
- #6: \n
- #7: \n
- #8: \n
- #9: shares, IaaS vs. PaaS\n
- #10: \n
- #11: \n
- #12: \n
- #13: \n
- #14: \n
- #15: \n
- #16: \n
- #17: fair share only what&#x2019;s IN THE QUEUE\nMinimum perf. + Burstability\n
- #18: \n
- #19: SMP applications assume they are running on physical CPUs\n
- #20: +vCPUs = harder to schedule\n
- #21: \n
- #22: any order, locks are held\ncpu stacking\nSMP kernel has concurrency, spinlocks, etc\n
- #23: make a good use of your CPU time (do not spend it with context switches)\n
- #24: hard to share I/O resources\nhard to fair share -> async\n
- #25: I/O is also heavily impacted by this\nisolate I/O components, attached resources\n\n
- #26: attachable resources\nbe prepared to scale-out\nvariability\n
- #27: \n
- #28: \n
- #29: data oriented discussions\n
- #30: \n
- #31: \n
- #32: \n
- #33: \n
- #34: \n
- #35: \n
- #36: \n
- #37: \n
- #38: \n
- #39: \n
- #40: \n
- #41: \n
- #42: \n
- #43: \n
- #44: internal, nagios\n
- #45: \n
- #46: \n
- #47: \n
- #48: help drive product decisions\n
- #49: absolute values are ok, be careful with bucket aggregators\n
- #50: URL or ALIAS DNS entries\n
- #51: \n
- #52: \n
- #53: \n