Minutes of Technical Community Discussion on February 05, 2015
Date and Time: UTC 14:00, Thursday Febuary 05, 2015Convenor: Bin Hu (AT&T)Participants:- Al Morton (AT&T)
- Andrea Pinnola (Telecom Italia)
- Bharat Mota (Freescale)
- Carlos Goncalves (NEC)
- Chris Adams (Itential)
- Christoph Meyer (Ericsson)
- Christopher Price (Ericsson)
- Hai Liu (Huawei)
- Haibo Zhang (Huawei)
- Joe Kidder (ENEA)
- Juha Oravainen (NSN)
- Linghui Zeng (Huawei)
- Luke Hinds (NSN)
- Manuel Rebellon (Sandvine)
- Maria Toeroe (Ericsson)
- Mario Torrecillas Rodriguez (ARM)
- Michael Bugenhagen (Century Link)
- Michael Godley (Intel)
- Nicolas St-Pierre (Sandvine)
- Prasad Gorja (Freescale)
- Qiao Fu (China Mobile)
- Raymond Nugent (Huawei)
- Stuart Mackie (Juniper)
- Tianran Zhou (Huawei)
- Ulrich Kleber (Huawei)
- Zhiqi Wang (Huawei)
- Discussion of Requirement project proposals
Zhiqi started with answering the question of adding this feature into Promise project and combine those 2 projects together. Zhiqi indicated that this proposal intends to enhance OpenStack's resource scheduler. Promise project intends to be based on StormForge. This is the key difference between this proposal and Promise project.
Zhiqi also indicated that both Ryota from Promise project and Zhipeng Huang from Copper project recognize the difference, and agree not to combine them together.
After clarification, group is satisfied with the proposal and agrees that this proposal is mature and ready for TSC review and approval. The consensus is to recommend it for creation review and approval in TSC.
Hai Liu introduced the update of proposal based on 4 action items from last meeting.
- The repo name is changed to "collector"
- The diagram is uploaded
- The relationship with Doctor project is clarified that predictor analyses data collected by the data collector and send notifications to Doctor module. The proposal has no direct relationship with Pharos and Copper
- In terms of the synergy with general mechanics of data collection, it was clarified that the data collector consists of Ceilometer and Monasca which can be extended to plugin some other data collector, e.g. zabbix, nagios, cacti.
It is further clarified that the actual work in data collector is the tool of hooking those plug-ins, and a module of metrics determining which data should be collected.
There was a question about if there is a plan to add an algorithm in Monasca / Ceilometer. It is clarified that the project just focuses on collecting data. Data processing is out of scope.
Another question was about the performance requirement when real-time data collection is performed. For example, how many data needs to be collected per second. It was clarified that both offline data collection and real-time data collection are supported. Offline data collection can train real-time behavior.
It was agreed that such performance requirement will be included in the deliverables of the project, like in big data, there is such type of requirement.
It was also agreed that disk space requirement will also be included in the deliverables of the project.
After discussion and clarification, group is satisfied with the proposal and agrees that this proposal is mature and ready for TSC review and approval. The consensus is to recommend it for creation review and approval in TSC.
The proposal author was not present. Uli questioned why it is still in the agenda of technical discussion since it is already in TSC agenda for creation review and approval.
Chris clarified that because someone asked him, so it was put in TSC agenda.
Bin reviewed prior meeting minutes and history of technical discussion. It was presented on Jan 15, and discussed on Jan 22. There was an action to update the proposal and clarify the scope and potential implementation path at Phase 2. The proposal author was absent on Jan 29 and today, so there was no further information about the status of action of updating the proposal and related clarification.
From the prior minutes, it doesn't show us a consensus on the maturity of this proposal.
Uli agreed to start an email thread to ask for further review and comment in order to address any potential concerns more effectively.
The meeting ended unexpectedly because another Test and Performance Meeting started 17 minutes before schedule. Starting another meeting automatically ended a previous meeting.
Postponed to next meeting.
- Discussion of Collaborative Development project proposals
Postponed to next meeting.
Postponed to next meeting.
Postponed to next meeting.
- Discussion of Integration and Testing project proposals:
None.
- Discussion of other projects
None.
Meeting adjourned
