Finding and fixing top performance issues with new relic rpm
7 likes2,147 views
This document summarizes Brian Doll's presentation on finding and fixing performance issues using New Relic RPM. It discusses the top 5 performance issues companies encounter: 1) repetitive work syndrome/insufficient caching, 2) web service dependencies causing domino effects, 3) insufficient system resources, 4) slow SQL queries, and 5) N+1 query problems. The presentation provides examples of each issue and explains how New Relic RPM can help companies monitor applications, troubleshoot issues, and optimize performance. It also discusses how RPM works and integrates with applications.
Finding and fixing top performance issues with new relic rpm
- 1. Finding and Fixing Performance Issues with New Relic RPM December 8th, 2010 @ 10AM PST Brian Doll - New Relic Application Performance Engineer
- 2. New Relic is the leading software as a service provider of application performance management tools. More than 6,000 companies use RPM, to manage more than 90,000 Java, .net, php, Ruby, or JRuby applications. RPM enables developers and operations teams to monitor, troubleshoot and optimize their web applications. Brian Doll Application Performance Engineering http://www.slideshare.net/briandoll
- 3. Monitors RPM Troubleshoots Tunes Web Applications In the Cloud or Datacenter NewRelic.com
- 4. Top 5 Performance Issues #5 N+1 Queries Death by a thousand paper cuts NewRelic.com
- 5. 70x per transaction! NewRelic.com
- 6. 1 parent + 15 children ___________________________________ 16 queries! NewRelic.com
- 7. Eager vs. Lazy Loading Eager vs. La zy 2 SQL queries vs. 16 SQL queries NewRelic.com
- 8. Top 5 Performance Issues #5 N+1 Queries #4 Slow SQL performance NewRelic.com
- 9. Transaction Traces Almost one full second! Missing index Explain Plans are Maps to Improving SQL Queries NewRelic.com
- 10. Top 5 Performance Issues #5 N+1 Queries #4 Slow SQL performance #3 Insufficient System Resources NewRelic.com
- 11. Waiting in line... NewRelic.com
- 12. How are we constrained? NewRelic.com
- 13. Scaling Silver Bullet? + ? + ? + ? NewRelic.com
- 14. Top 5 Performance Issues #5 N+1 Queries #4 Slow SQL performance #3 Insuf鍖cient System Resources #2 Web Service Domino Effect NewRelic.com
- 15. Uh oh... NewRelic.com
- 16. Search service unavailable leads to... Unhappy customers leads to... Lost sales !%&'()* !"#$ !"#$ !"#$ !"#$ ! "#$ /b NewRelic.com
- 17. Circuit Cascading Breakers Failures NewRelic.com
- 18. Top 5 Performance Issues #5 N+1 Queries #4 Slow SQL performance #3 Insuf鍖cient System Resources #2 Web Service Domino Effect #1 Repetitive Work Syndrome or Cache as cache can NewRelic.com
- 19. Slow Lots of work NewRelic.com
- 20. Caching is easy. Cache invalidation is not. NewRelic.com
- 21. How RPM Works a! a! a! a! Https Applications New Relic Service Https 犢 Secure, reliable service 犢 No private user data 犢 View app performance from anywhere App Team Supports NewRelic.com
- 22. + Faster apps = Happy customers NewRelic.com
- 23. Installs in 2 minutes Supports Joyent SmartMachines Java Ruby Special Offer: PHP 15% discount for Joyent customers NewRelic.com
- 24. Q&A
- 25. Questions? Live Q&A during the webinar http://support.newrelic.com/ Email: support@newrelic.com Twitter: @NewRelic IRC: #NewRelic on Freenode NewRelic.com