Network Caching and Programmable Hardware Insights
Discover the benefits of in-network caching, the role of caching in reducing latency levels, and how caching aids in load balancing within backend systems. Learn about programmable network hardware running OS for off-data-path computation, packet mirroring for network analysis, and the usefulness of recirculation in handling limited stages. Dive into a breadth of experiments evaluating system design limits and future potentials. Explore NetCache for workloads, multi-rack scenarios, and testing methodologies. Uncover insights from SwitchKV and IncBricks for network traffic direction and data caching using processing units.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
In-network caching CSE 552 preparation for reading
Caching with fast memory decreases latency Level-i cache Level i+1 storage or cache
Caching helps with load balance Cache of size O(n log n) n = number of backends Backend 0000 3FFF Backend 4000 7FFF Backend 8000 BFFF Backend C000 FFFF
Programmable network hardware Switch runs an OS, which can be used for off-data-path computation Some modern switches are programmable, e.g., in the P4 language
Packet mirroring Packet mirroring is typically used for network analysis: Send packet where it s supposed to go, and to an analyzer machine
Recirculation Recirculation is useful when the limited stages aren t sufficient
Note the breadth of experiments They evaluate beyond the limits of the system s design range NetCache is designed for read-mostly workloads, but evaluation includes heavy-write workloads They even evaluate the potential for future work NetCache only operates in a rack, but they simulate a multi-rack scenario They use a snake test to evaluate as if they had more computers Unfortunately, they don t quantitatively compare to other systems SwitchKV: Program the network to direct traffic to caches IncBricks: Use network processing units, not switches, to cache data
