Forum Replies Created
-
Posts
-
Hello Mazahir,
This issue was escalated and on checking, it appears all the slices have been “closed”. See attached output from management-cli.
Please could you confirm this?
Regards,
Acheme
Hi Bohdan,
Yes, with FABlib you can create your slice with the VMs and the networking needed. Also your script needs to have a time dependent function that performs the test at the intervals you specify using addresses of the nodes you created earlier. You can save the results in the FABRIC nodes or scp them to your local computer as well.
I am not sure there is such an example script, if that is what you are asking though.
The maximum interface Bandwidth on the testbed is 100G (with ConnectX-6 SmartNics). How much of this is achievable depends on the end-to-end link capacity and background network traffic at the time. Also, you might need some CPU pinning and NUMA tuning to achieve high throughput in iperf tests, see example: https://github.com/fabric-testbed/jupyter-examples/blob/main/fabric_examples/complex_recipes/iPerf3/iperf3_optimized_smart_nic.ipynb
Hello Luca,
Can you confirm you have routes (most likely static route) on node 2 and 3 to forward traffic to node4 and 1 respectively?
Also, could you share your notebook so I can try to replicate the issue.
Regards,
Acheme
Thank you Ben.
I also got information that if you don’t find a link mapped on the dashboard, it is likely the sites involved are not directly connected.
I configured the IP address as it was before it crashed but cannot ping the other nodes using the dataplane. Is there something else I should do?
