QoS metrics are, in principal, any aspect of the storage service that researchers care about when interacting with it. QoS metrics definitions are intended to be device independent (they may be used to describe different storage services, based on distinct technologies), transport independent (they do not depend on how the users interact with a storage service), and time independent (they should always apply to all data stored with some given QoS).
A common set of QoS metrics will allow users to compare different services and choose the most appropriate for their use-case. It also will allow brokering services to aggregate storage to provide a synthetic or aggregated storage service with well-defined behaviour.
Data lifecycle activity are transitions that researchers can perform and that they know in advance will be needed. The trigger may be simply elapsed time, or may be something related to the data.
A common data lifecycle language will allow researchers to off-load certain responsibilities for managing their data to storage systems. Since such storage systems are often common to many research groups, this concentration of responsibility reduces the effort required to maintain data.
2. Work on finalising Case Statement (all), as necessary,
3. Presentation of existing QoS work (named speakers),
4. Decision on which metrics to focus (all),
5. Initial set of definitions (all).
Ideally participants should join the QoS-DataLC mailing list and introduce themselves prior to attending the meeting.