Prediction project meeting

Please join my meeting from your computer, tablet or smartphone.https://global.gotomeeting.com/join/941554773

You can also dial in using your phone.

United States (Long distance): +1 (571) 317-3129

Access Code: 941-554-773

1. Use case discussion

1. Currently we found some confused points in our use cases, so we discussed to discovery new ones. We reconsidered the time series and talked about the Fault tree analysis.

1. After discussion, we came to an agreement that BP work is not at the high priority. So we will pending it.

2. For the development and demo plan, we have finished basic data collection. Next we will mainly working on prediction algorithm, we can start from some simple use cases. At the end of the demo plan, we will develop the prediction dashboard, currently one choice is to use python django.

1. Jira update. 11 of 13 Jiras have been added by Yifei since last week and the other 2 of 13 were abandoned.

2. Jira discussion. 1 Jira was closed by Linghui.

1. Work plan discussion for Milestone D of B release.

2. Jira update. 13 Jiras will be added by Yifei after discussion.

1. Yifei Xue introduced VMI technology.

2. Yifei thought that the VMI would be not appropriate for our project as the performance overhead. Hai liu said that the support of monitoring for third part APP seemed not good.

1. Linghui Zeng and Yifei Xue introduced the agentless monitoring technology.

2. Yuanxun Gu asked the basic information about the prediction project.

1. Linghui Zeng introduced two use cases of the project. If possible, some more reasonable use cases will be added.

2. Hai Liu explained four gaps between our requirements and OpenStack. Two of them have been descript as BPs for Monasca in Launchpad.

3. For the questions about Release B plan, Yifei Xue said that Prediction team will have a conversation with the Monasca team to make these two BPs approved, then code will be contributed. In Release B, Prediction team will finish the requirements doc and complete the BP development for data collection. Predictor will be considered in Release C.