Riverbed - Enforcing User-defined Privacy Constraints in Distributed Web Services
Riverbed: A framework for privacy-respecting web services, allowing users to define data processing constraints. Uses remote attestation and IFC to enforce policies without requiring code modifications.
Shoal - A Network Architecture for Disaggregated Racks
Innovative network architecture for disaggregated racks using fast circuit switches, offering high density and performance with lower power consumption. Addresses unique challenges in connecting dense clusters of compute, memory, and storage blades.
Understanding Lifecycle Management Complexity of Datacenter Topologies
Explores lifecycle management complexity of datacenter topologies, comparing Clos and expander graphs. Introduces FatClique, a new topology combining wiring simplicity and expandability, with insights on deployment and expansion costs.
Monoxide - Scale Out Blockchains with Asynchronous Consensus Zones
Explores Asynchronous Consensus Zones to scale blockchain systems linearly, introducing eventual atomicity and Chu-ko-nu mining for improved throughput and security without compromising decentralization.
NetBouncer - Active Device and Link Failure Localization in Data Center Networks
Active probing system for accurate failure localization in data centers. Detects device and link failures, integrates domain knowledge with machine learning for high accuracy and resilience to data inconsistency.
Hyperscan - A Fast Multi-pattern Regex Matcher for Modern CPUs
Efficient regular expression matching for network security using graph decomposition and SIMD acceleration, improving performance of applications like Snort by 8.7x on real traffic.
Solving All the Problems with systemd
Explore systemd's advanced features for flexible service management, combining traditional and container-like approaches for optimal system control and security.
SRE and DevOps at a Startup
Exploring SRE and DevOps implementation strategies for startups, comparing centralized and distributed team approaches, and discussing challenges and lessons learned.
Deploying SRE Training Best Practices to Production - What We Learned
Explore best practices for SRE training, including sequential learning, hands-on experiences, and continuous education. Learn from Google's insights on building effective SRE teams and fostering a culture of reliability.
Canarying Well - Lessons Learned from Canarying Large Populations
Explore canarying best practices, pitfalls, and strategies for safe production changes. Learn to balance priorities, handle diverse scenarios, and implement effective canary processes in software deployment.
SRE Theory vs. Practice - A Song of Ice and TireFire
Humorous talk contrasting idealized SRE practices with real-world challenges. Speakers debunk perfect environments, offering practical insights and relatable experiences for SRE professionals.
Availability, Latency, and Cost - Withstanding Regional Outages
Explore Netflix's multi-region strategy for improved availability and latency, including algebraic models, incident management, and design considerations for efficient failovers and user steering.
The Guard's Dilemma - Efficient Code-Reuse Attacks Against Intel SGX
Explores novel exploitation techniques against Intel SGX, demonstrating vulnerabilities in enclave code without requiring kernel privileges or crashes, even with existing randomization approaches.
Unveiling and Quantifying Facebook Exploitation of Sensitive Personal Data for Advertising Purposes
Unveils Facebook's exploitation of sensitive user data for advertising, quantifying its prevalence and discussing privacy implications under GDPR. Proposes a browser extension for user awareness.
Efail - Breaking S-MIME and OpenPGP Email Encryption Using Exfiltration Channels
Uncovers vulnerabilities in OpenPGP and S/MIME email encryption standards, demonstrating how malleability gadgets can be exploited to reveal encrypted content through exfiltration channels in various email clients.
GAZELLE - A Low Latency Framework for Secure Neural Network Inference
Efficient framework for secure neural network inference combining homomorphic encryption and two-party computation, outperforming existing systems in speed while preserving privacy for cloud-based machine learning.
MoonShine - Optimizing OS Fuzzer Seed Selection with Trace Distillation
Optimizing OS fuzzer seed selection through trace distillation, improving code coverage and vulnerability detection in Linux kernel testing.
FUZE - Towards Facilitating Exploit Generation for Kernel Use-After-Free Vulnerabilities
Explore automated exploit generation for kernel Use-After-Free vulnerabilities using FUZE framework, enhancing security analysis efficiency and exploitability evaluation for complex OS kernels.
QSYM - A Practical Concolic Execution Engine Tailored for Hybrid Fuzzing
Explores QSYM, a fast concolic execution engine for hybrid fuzzing, integrating symbolic emulation with native execution to find bugs in complex software more efficiently than traditional methods.
Malicious Management Unit - Why Stopping Cache Attacks in Software is Harder Than You Think
Explores indirect cache attacks bypassing software-based defenses by exploiting MMU operations, challenging existing security measures and revealing vulnerabilities in cache isolation techniques.
NetHide - Secure and Practical Network Topology Obfuscation
NetHide: A framework for network topology obfuscation that mitigates link-flooding attacks while preserving debugging capabilities. Leverages programmable network devices to modify path tracing probes at line-rate, balancing security and usability.