The QoS White Paper
The Working Group has published a white paper describing
QoS and its potential within WLCG. The document is intended to frame and to stimulate the necessary debate on how the community can progress in realising the benefits of
QoS.
The latest released version of the white paper is
v1.0.0.
The advanced draft of the next version is available on
Google docs.
The expected lifecycle of the document is as follows
- Initial publication of v1 - Jan 2020
- Distribution to experiments and other stakeholders with request for a response
- Discussion in the group and perhaps at a dedicated meeting
- Publication of a final version incorporating the results of the consultation
Request for feedback
The WG actively solicits responses from stakeholders to this document.
These should be sent to
wlcg-doma@cernNOSPAMPLEASE.ch.
In particular, considerations of what actions involving experiments and sites could be taken.
- Attempt a more granular QoS classification and workflow mapping.
- Propose a set of new QoS classes
- Invite sites to begin a classification of their current offerings
- Based on a set of initial propositions
- Sites could also add any they provide (or are interested in providing) but do not figure in the list
- Could include hints on cost and reliability (e.g. indications of replication level)
- A mechanism for exposing these must be found (CRIC?)
- Identify where workflows can be adapted to be less demanding on storage.
- Invite sites to report on current relevant directions
- Media diversity
- Redundancy layer
- In particular, introducing this into existing systems
- Attempt the "JBOD experiment" in conjunction with an experiment.
- Automate recovery from data loss to allow reliability requirements to be relaxed where appropriate.
- Remove all redundancy (gaining ~15% capacity) and work on handling data loss gracefully
- Ask sites to provide new experimental storage areas with novel QoS features
- Trigger exploration of what adaptations are needed over the stack
- Experiment/Site combinations could be encouraged to report.
- Co-evolve workflows and storage configurations - the Atlas carousel R&D is an example of this.