Skip to end of metadata
Go to start of metadata
Attendees
Sridhar Rao
Shreya Gupta
Shubhank Saxena
Al Morton
Quang-Huy Nguyen
Luc Provoost
Sl. No | Topic | Presenters | Notes |
---|
1 | Update from Interns |
| Shreya and Shubham - Understand Helm well now. Deploying sample helm charts. Single and multi-pod deployment. Testing on Pod18.
- Post-Deployment access. Exposing the pod. *SSH Tunnel*.
- Trying trafficgen pod and L2fwd Pod - Helm-chart for rapid-prox pod and l2fwd.
ToDo: Single pod with 2 containers (rapid and l2fwd), using helm. |
2 | Prox Results Discussion |
| - Run-1: bucket_size_exp = 24, maxz = 5000, latency_cores: 14-16 (if the number of packets lost during the run is higher than maxz, DO NOT rerun)
- Run-2: bucket_size_exp = 14, maxz = 5, latency_cores: 14-16
- Run-3: maxz = 50000, latency_cores: 14
Check for the cause for tx-error. Tests are unidirectional only. To make it bidirectional, we have 2 options: - Create another set of gencores.
- Use existing set of gencores, but add another task.
Luc Provoost will guide us which one of the above 2 is the best. |
3 | Prox-TGen Pod Usage Update |
| - exclude sleep command.
- crashing when gencores is more than 1 (segmentation fault).
Yury (Intel) is helping in debugging this issue. |
4 | VineProx Update | | Fully functional - Generate the config and run test. |
|
|
|
|