[Hps] pass-0 reconstructed data sets for engineering run

[McCormick, Jeremy I.]

Or it could be made meta data in the data catalog by file….

On Dec 29, 2014, at 6:28 PM, Graham, Mathew Thomas <mgraham at slac.stanford.edu<mailto:mgraham at slac.stanford.edu>> wrote:


Hi All,

The “0th pass” of the evio to reconstructed lcio is done***, you can see a summary of what exactly was run at this page:

https://confluence.slac.stanford.edu/pages/viewpage.action?pageId=184716812

…I used an untagged release; basically the HEAD on 12/25.  Included in the lcio:

— trigger info
— raw hits
— gain corrected hits
— ECal clusters for 3 clusters: EcalClusterer, EcalClustererIC, GTPOnlineClusterer

 Running the entire set of runs takes ~1 day (or so, we had some database issues that made it take longer but they have been fixed) and the files take up roughly 5TB.

In the next few days I’ll run the data quality and the DST maker.  When we get back from break we’ll discuss what we should add/subtract/change for the next reconstruction pass.

Thanks, Matt

*** only 6 out of 1493 jobs have failed at this point; I’m looking in to the stragglers.
_______________________________________________
Hps mailing list
Hps at jlab.org<mailto:Hps at jlab.org>
https://mailman.jlab.org/mailman/listinfo/hps

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/hps/attachments/20141230/22b64e75/attachment.html