[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: test plan today
Hi Alex,
I recommend varying the EXECTIME. The ones I made are just suggestions.
Also, if youd like to reduce the queries in size you can freely vary
tagrand which is a random number tag which runs from 0-2000
for example if you would prefer to get
108 events instead of 217 events for the 2nd query
modify it by making
glb_trk_tot>5000&mean_pt>0.5&tagrand>1000.
Also, if you would like to select 1% of the events in the database
randomly distributed through the files (not really since the data bases
have different numbers of events) just ask for tagrand<20 as your query.
david
> Hi,
> starting at little after 2pm (EST) today,
> I'm going to run the following 5 queries in the order ...
> Any suggestions welcome...
>
> mean_eta<-0.2 EXECTIME: 200
> 100 events in 55 files.
>
> glb_trk_tot>5000&mean_pt>0.5 EXECTIME: 5000
> 217 events in 80 files
>
> glb_trk_tot>0&glb_trk_tot<10&n_vert_total<3 EXECTIME: 10
> 40 events in 29 files
>
> mean_eta>0.2 EXECTIME: 200
> 113 events in 55 files.
>
> glb_trk_tot<30 EXECTIME: 100
> 285 events in 71 files.
>
> First, I'm going to run each query with SII w/o Objectivity involved,
> but with HPSS with cmanager.old. I'm going to clean the destination
> (staging) directory everytime so that CM can actually cache all the
> files that it needs.
>
> Second, I'm going to run with the user code, and someone should let me
> know where the
> user code is, and how to run it with a query...
>
> That'll take a whole day today, and from tomorrow, Henrik, Luis and Arie
> will take over
> the test at BNL.
>
> Thanks alot...
> --Alex
>
>
> --
> Alex Sim
> MS 50B-3238 http://sim.lbl.gov
> DM, NERSC, LBNL ASim@lbl.gov
> 1 Cyclotron Road office: (510) 495-2290
> Berkeley, CA 94720 FAX: (510) 486-4004
>