User Interface to TAG Data Services

Note that this is a list of desireables which will not necessarily be implemented in a prototype but that the prototype should not rule out implementing later.

Developers

Jack, Marco, Jerry

Description

Skimming Services

Information a first pass user will likely have
  1. a 'dataset' of interest
  2. a release version or run of production needed
  3. a qualifier on the calibrations used
  4. a set of qualifiers (time, quality, trigger, etc.) which can be turned into a query

The information for (1), (2), (3) will apply to an entire collection, while (4) will apply to the elements of a collection.

Information that the user might request be applied to the output
  • a place to deliver the output
  • a name for the output
  • a partitioning model, i.e. group by
    • input dataset
    • number of events
    • input files
    • running periods
    • ...
  • the name of a new collection which he wants to have visible in the same way as the production collections/TAGs

Information a second pass user might have
  • the name of a previous job/task for which he wants to repeat the same query
  • the name of the output of a previous selection

Additional (Non-Skimming) Services

Additional questions a user might ask which in addition to skimming operations:
  • What collections are available to query?
  • What metadata is available on those collections?
  • How many events satisfy a query?
  • Which DDM datasets contain the results of a query?
  • How many new events are available since I made the original query which built my skim? Can I simply update my existing collection/dataset with these new events?
  • Do I get different events if I do the same query on a collection created with a different calibration/software version?
  • Is this TAG file dataset 'equivalent' (i.e. would return the same query results) to a relational TAG database?
  • If I wanted to run this at another site, is there an 'equivalent' TAG database available there?
  • Can I make a selection on multiple partial TAG databases?
  • Can I get a consistent subset of data, TAG, conditions which I can export to my Tier 3+?

Site Administration

Capabilities a site administrator might want to have
  • I have a good connection to the Tier 1, can I forward a query to a remote server?
  • I need to bring down a disk or database for maintenance. Can I change a configuration to use a replica TAG database?
  • I want to have a section of the database that is read-only and another where users can write.
  • I want to have different visibility for different sections of the database, i.e. local, Tier2, remote, ...
  • I want to have failover options which are under local control.

Dependencies

  • Most everything

-- JackCranshaw - 15 Aug 2006
Topic revision: r5 - 06 Nov 2006, RobGardner
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback