Best UADK Acceleration Practice on Ceph Storage
summary
UADK integration offloads Ceph encryption and compression into hardware using IOMMU + shared virtual address to avoid memory copies and enforce device-boundary security, exposing an OpenSSL engine and a Ceph crypto/compressor plugin (a ~500-line patch) with CPU fallback. Offloading AES and MD5 cuts CPU use by about 50%, hardware compression plus offload reduces CPU similarly and increases throughput roughly 60–68%. Works on BlueStore and RGW with no major Ceph code changes.
AI-written summary · about
chapters (7)
- 0:00Introduction to UADK
- 1:26Ceph Storage Data Process
- 3:54Challenges and UADK Solution
- 6:05Security Offload Practice
- 8:13Data Compression Offload
- 10:26Other Optimizations
- 11:32Future Work and Conclusion
related talks
27mEncrypting Data in Ceph: An Overview and Opportunities
47mLinux Cryptographic Acceleration on an i.MX6
30mAccelerate TLS with DPDK
27mSpeeding Up Linux Disk Encryption
29mSecuring Data in Motion and at Rest: Quantifying End-to-End Encryption Perfo...
35mData Security and Storage Hardening in Rook and Ceph
36mCryptodev and Security Library Updates in DPDK
30mData Security: Theoretical and Real World Approache...