LHCb Full Experiment System Test (FEST'09)

Aim

FEST'09 is meant at testing the full processing chain from the input to the HLT farm to the stripping/streaming and distribution of DSTs. It was discussed and presented in several places:

FEST planning

It is foreseen that FEST tests will be run for a duration of one week, followed by a few weeks for taking lessons of the results and preparing the next week. Objectives of increasing complexity will be introduced. FEST'09 should run at least until LHCb starts again data taking from the detector.

The planning of FEST'09 weeks is below:

  • 26-30 January
  • 2-6 March

Preparation for the January FEST week last update 21.01.09)

  1. Agree on a namespace for files (c.f. this page). I suggest to use the same convention as for CCRC08, i.e. we use the standard namespace, just replace the <year> with "FEST09". Same for the "ConfigVersion" (i.e. the "partition") : "FEST09". Note the addition of <stream> after "RAW" as agreed in November. The <stream> for all events should be chosen (I propose "Full")
  2. Install DIRAC v4r4 on the Online system (Andrew to pass over to Joel, Rashid should also follow)
  3. Check the XML generated from runDB after recent modifications (Daniel, Jean-Christophe, Zoltan)
  4. Trigger a few transfers manually to verify that the full loop works: transfer, integrity check (file migration), feedback to online, LFC registration, run registration in BK, gotReplica flag in BK set when transfer+migration successful. Verify that files do appear in the BK, create an interactive job to read one, getting the options from BK_GUI (name?)
  5. Verify registration in transferDB, and transfer to Tier1s (set low limit on number of files, as rate will be very low) (Andrew, Rashid)
  6. Prepare a reconstruction workflow (Stuart), a processing transformation (Stuart) with regular expression matching to start with (use run number?). Launch reconstruction production in manual first, in automatic then when/if many files. RDST should be stored locally on LHCb_RDST space.
  7. Prepare an XpressLine processing production. Should we keep the RDST or only histograms? Note: saving histograms should be agreed with Patrick K and implemented in the workflow (which storage: LHCb_LOG? space token (T0D1) in Castor?).

-- PhilippeCharpentier - 21 Jan 2009

Edit | Attach | Watch | Print version | History: r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r1 - 2009-01-21 - PhilippeCharpentier
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    LHCb All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback