Talks
Speakers
Events
Topics
Sign in
Home
Talks
Speakers
Events
Topics
Leaderboard
Use
Analytics
Sign in
Suggest modification to this talk
Title
Description
@techbint Whenever something has gone seriously wrong for me in production caching has been at the root of it. Live Q&A with Julian Assange almost brought the site down? Caching fail. Servers can't cope with traffic about a new woolly rat? Caching fail. Half the site is showing sorry pages? Caching fail. This talk will use these disasters to explain why the simplest caching is always the best. Talk given at GORUCO 2015: http://goruco.com
Date
Summarized using AI?
If this talk's summary was generated by AI, please check this box. A "Summarized using AI" badge will be displayed in the summary tab to indicate that the summary was generated using AI.
Show "Summarized using AI" badge on summary page
Summary
Markdown supported
In this talk titled "Great Caching Disasters!" presented by Lisa van Gelder at GoRuCo 2015, the focus is on the significant challenges and failures experienced due to caching issues, particularly during her time at the Guardian. The speaker shares various anecdotes to illustrate how simplistic caching solutions are often the most effective. **Key Points Discussed:** - **Introduction to Caching Problems:** - Van Gelder experienced downtime at the Guardian due to faulty caching practices, notably during peak traffic hours. - **Impact of Poll Submissions:** - Investigating logs revealed that cache clearances coincided with poll submissions, leading to service disruptions. - **Lessons Learned:** - Importance of closely monitoring caching mechanisms and avoiding manual cache clearance options due to their propensity for misuse. - **Cache Duration:** - A three-day cache duration led to issues when articles were deleted, reinforcing the lesson that shorter caching can be more effective. - **Case Study – Live Q&A with Julian Assange:** - The Guardian's comment system faltered under load despite having a short HTML fragment cache, highlighting the risks of integrating processing instructions with cached content. - **'Woolly Rat' Phenomenon:** - Van Gelder refers to a viral article that caused unexpected traffic, illustrating the necessity for robust caching processes that respond automatically to traffic spikes. - **Final Recommendations:** - Simplify caching practices, monitor performance closely, and avoid complex systems unless caching is the core business focus. **Conclusion:** Van Gelder emphasizes the importance of simplification and vigilance in caching management. By adopting practices like maintaining a one-minute cache and avoiding manual interventions, organizations can significantly reduce the likelihood of caching disasters. The talk serves as a valuable lesson for engineers and developers in managing caching systems effectively, particularly during high-traffic events. The speaker concludes by advising against elaborate caching systems unless specifically needed, and encourages utilizing Content Delivery Networks (CDNs) for effective caching solutions, leaving complex setups to the experts.
Suggest modifications
Cancel
