ns3-rdma
ns3-rdma copied to clipboard
How to compile and run with WAF under Ubuntu?
Hello, I learned a lot from your project. Now I want to ask a question about WAF compilation. I have run successfully under Windows, now I want to merge it into the ns3.29 project under ubuntu. What should I do? Look forward to your reply!
Find wscript in each code folder, e.g., https://github.com/bobzhuyb/ns3-rdma/blob/master/src/point-to-point/wscript and add the source files that were not in the original code.
Merging into ns3.29 is non-trivial -- a lot of APIs changed in ns-3 between versions. Good luck with that.
Thank you for your reply! Some time ago, I tried to debug NS3.29 according to this method. There are still some problems. If it can't be solved, I want to continue to ask you. Thanks again!
Hello! Can the slow-recevier symptom mentioned in the paper be reproduced in the simulator? Can the dynamic allocation buffer scheme in the solution be implemented through parameter adjustment?
The slow receiver is not implemented in this simulator. However, you can always implement something at the receiver side to slow down the receiving, or to send out PFC pauses.
"Dynamic allocation buffer scheme" -- if you are referrint to the dynamic PFC threshold, it has been implemented. Check config.txt. There is a knob to turn this on and off.
Hello! I have been studying the literature on RDMA recently. Among them, there is a document named "Memory Efficient Loss Recovery for hardware-based Transport in Datacenter", which is written by Yuanwei Lu, Peng Cheng and other scholars. It is mentioned in the literature that "MELO has been integrated into DCQCN in NS3". I have two questions to ask you, thank you. A. Where is melo's code slice implemented on DCQCN? B. How is memory resource tracing implemented in NS3? (or in this case) These two points are very important to me. I am looking forward to your reply.
You'd better ask the authors. I believe they cloned this project and made their modifications in their own repo.
Thanks for your reply, what should I do if I try to track memory data for your project? Thank you!
Do you mean the memory that would be used by the simulated end nodes? You can track this value https://github.com/bobzhuyb/ns3-rdma/blob/e8a27d02bfe5ef0aec189c7d1a401547dbd8fbad/src/applications/model/udp-client.cc#L173
Thank you very much for your reply!
Hi! Thank you for always keeping in touch with me. At the moment, I want to try to write a simple iconic message similar to the pfc pause frame to inform the upstream device port to do something, but there are some doubts about the nature of the pfc pause frame. Is the pfc pause frame a packet or a stream identifier? How does it do reverse transmission? Looking forward to your reply!
PFC frames are packets generated and consumed by qbb-net-device
https://github.com/bobzhuyb/ns3-rdma/blob/master/src/point-to-point/model/qbb-net-device.cc#L965
Hi,Yibo! I want to know the source code location of the tracking variable. In addition, I want to get variables such as the buffer queue length change, latency, and throughput in the same way. What should I do? Looking forward to your reply!
Sorry, I don't seem to understand where the simulation duration is set. Can you tell me where the 1 second is set,please? Thank you!
Thank you for your reply! Some time ago, I tried to debug NS3.29 according to this method. There are still some problems. If it can't be solved, I want to continue to ask you. Thanks again!
Hello, I want to ask, how is your merger now?