CDN Monitoring & Troubleshooting 5 — Questions and Answers
Question 1: What is 'cache warming' in the context of CDN monitoring, and when is it critical?
- Increasing server room temperature for optimal hardware performance
- Pre-populating edge caches with content before a traffic event to prevent origin overload on cold caches (Correct answer)
- Monitoring cache temperature metrics in dashboard
- Gradually increasing CDN subscription tier
Correct answer: Pre-populating edge caches with content before a traffic event to prevent origin overload on cold caches
Cache warming involves pre-fetching content to edge nodes before high-traffic events (product launches, live streams) so initial requests are served from cache rather than overwhelming the origin.
Question 2: A CDN's access logs show requests with 'cf-ray' headers but users in Asia report high latency. What does this indicate about routing?
- The CDN is functioning correctly and latency is a client-side issue
- Users may be routed to a distant POP due to BGP routing anomalies or missing regional POP coverage (Correct answer)
- The cf-ray header indicates a caching error
- SSL certificates are expired in Asian POPs
Correct answer: Users may be routed to a distant POP due to BGP routing anomalies or missing regional POP coverage
CF-Ray (Cloudflare's request ID) encodes the POP identifier; if Asian users are connecting to non-Asian POPs despite Anycast routing, BGP path issues or POP capacity problems are likely.
Question 3: Which monitoring approach helps detect 'cache poisoning' attacks on a CDN?
- Monitoring total bandwidth consumption only
- Alerting on unexpected changes in cached object sizes or content hashes for static assets (Correct answer)
- Tracking DNS query volumes per zone
- Monitoring SSL certificate renewal dates
Correct answer: Alerting on unexpected changes in cached object sizes or content hashes for static assets
Cache poisoning injects malicious content into CDN caches; detecting unexpected changes in cached object content hashes or sizes for known static assets can reveal such attacks.
Question 4: A company's CDN shows normal global performance metrics but a specific enterprise client on a private MPLS network reports 100% cache misses. What is the most likely cause?
- The client's content is not eligible for caching
- The client's egress IP resolves to a CDN POP that has not yet replicated the cached content, or their proxy strips cache-related headers (Correct answer)
- The CDN's global TTL is set too low
- The client is using IPv6 which bypasses CDN caching
Correct answer: The client's egress IP resolves to a CDN POP that has not yet replicated the cached content, or their proxy strips cache-related headers
Enterprise clients behind proxies or MPLS networks may egress through IPs routed to less-populated CDN POPs, or their network proxy may strip Vary/Cache-Control headers affecting CDN cache decisions.
Question 5: What does monitoring the CDN's 'origin connection time' metric help diagnose?
- End-user download speed
- Latency between the CDN edge and origin server, useful for detecting origin network or geographic placement issues (Correct answer)
- Number of concurrent CDN users
- SSL certificate chain validation duration
Correct answer: Latency between the CDN edge and origin server, useful for detecting origin network or geographic placement issues
Origin connection time measures how long CDN edge nodes take to establish connections to the origin, identifying whether origin is geographically far from edges or experiencing network congestion.
Question 6: During CDN troubleshooting, a developer uses curl -I to check headers and sees 'Age: 0' on every request. What does this confirm?
- The content has been cached for 0 seconds and will expire immediately
- The CDN is not caching the response (every request results in a fresh origin fetch) (Correct answer)
- The origin server clock is misconfigured
- The CDN TTL is set to exactly 1 second
Correct answer: The CDN is not caching the response (every request results in a fresh origin fetch)
Age: 0 on every request means the CDN is fetching fresh content from origin for each request rather than serving from cache, indicating a caching configuration problem.
Question 7: A CDN operator wants to identify which content types consume the most origin bandwidth. Which report or metric should they pull?
- Total CDN request count by status code
- Origin traffic breakdown by MIME type or file extension with byte transfer volumes (Correct answer)
- Client device type distribution
- SSL handshake failure rate by cipher suite
Correct answer: Origin traffic breakdown by MIME type or file extension with byte transfer volumes
Breaking down origin bandwidth by MIME type or file extension reveals which content categories (video, images, JS) are generating the most origin fetches for cache miss optimization.
CDN Monitoring & Troubleshooting 5
What is 'cache warming' in the context of CDN monitoring, and when is it critical?