TY - GEN
T1 - Optimizing overlay-based virtual networking through optimistic interrupts and cut-through forwarding
AU - Cui, Zheng
AU - Xia, Lei
AU - Bridges, Patrick G.
AU - Dinda, Peter A
AU - Lange, John R.
N1 - Copyright:
Copyright 2013 Elsevier B.V., All rights reserved.
PY - 2012
Y1 - 2012
N2 - Overlay-based virtual networking provides a powerful model for realizing virtual distributed and parallel computing systems with strong isolation, portability, and recoverability properties. However, in extremely high throughput and low latency networks, such overlays can suffer from bandwidth and latency limitations, which is of particular concern if we want to apply the model in HPC environments. Through careful study of an existing very high performance overlay-based virtual network system, we have identified two core issues limiting performance: delayed and/or excessive virtual interrupt delivery into guests, and copies between host and guest data buffers done during encapsulation. We respond with two novel optimizations: optimistic, timer-free virtual interrupt injection, and zero-copy cut-through data forwarding. These optimizations improve the latency and bandwidth of the overlay network on 10 Gbps interconnects, resulting in near-native performance for a wide range of microbenchmarks and MPI application benchmarks.
AB - Overlay-based virtual networking provides a powerful model for realizing virtual distributed and parallel computing systems with strong isolation, portability, and recoverability properties. However, in extremely high throughput and low latency networks, such overlays can suffer from bandwidth and latency limitations, which is of particular concern if we want to apply the model in HPC environments. Through careful study of an existing very high performance overlay-based virtual network system, we have identified two core issues limiting performance: delayed and/or excessive virtual interrupt delivery into guests, and copies between host and guest data buffers done during encapsulation. We respond with two novel optimizations: optimistic, timer-free virtual interrupt injection, and zero-copy cut-through data forwarding. These optimizations improve the latency and bandwidth of the overlay network on 10 Gbps interconnects, resulting in near-native performance for a wide range of microbenchmarks and MPI application benchmarks.
UR - http://www.scopus.com/inward/record.url?scp=84877686966&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84877686966&partnerID=8YFLogxK
U2 - 10.1109/SC.2012.88
DO - 10.1109/SC.2012.88
M3 - Conference contribution
AN - SCOPUS:84877686966
SN - 9781467308069
T3 - International Conference for High Performance Computing, Networking, Storage and Analysis, SC
BT - 2012 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2012
T2 - 2012 24th International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2012
Y2 - 10 November 2012 through 16 November 2012
ER -