CDNs’ Dark Side: Security Problems in CDN-to-Origin Connections
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Content Delivery Networks (CDNs) play a vital role in today’s Internet ecosystem. To reduce the latency of loading a website’s content, CDNs deploy edge servers in different geographic locations. CDN providers also offer important security features including protection against Denial of Service (DoS) attacks, Web Application Firewalls (WAFs), and recently, issuing and managing certificates for their customers. Many popular websites use CDNs to benefit from both the security and the performance advantages. For HTTPS websites, Transport Layer Security (TLS) security choices may differ in the connections between end-users and a CDN (front-end or user-to-CDN), and between the CDN and the origin server (back-end or CDN-to-Origin). Modern browsers can stop/warn users if weak or insecure TLS/HTTPS options are used in the front-end connections. However, such problems in the back-end connections are not visible to browsers or end-users, and lead to serious security issues (e.g., not validating the certificate can lead to MitM attacks). In this article, we primarily analyze TLS/HTTPS security issues in the back-end communication; such issues include inadequate certificate validation and support for vulnerable TLS configurations. We develop a test framework and investigate the back-end connection of 14 leading CDNs (including Cloudflare, Microsoft Azure, Amazon, and Fastly), where we could create an account. Surprisingly, for all the 14 CDNs, we found that the back-end TLS connections are vulnerable to security issues prevented/warned by modern browsers; examples include failing to validate the origin server’s certificate, and using insecure cipher suites such as RC4, MD5, SHA-1, and even allowing plain HTTP connections to the origin. We also identified 168,795 websites in the Alexa top 1 million that are potentially vulnerable to Man-in-the-Middle (MitM) attacks in their back-end connections regardless of the origin/CDN configurations chosen by the origin owner.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it