[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #NTD-472266]: Excessive loading of data in bundles



> Yuan:
> 
> I had a chance to try the nightly build today. I'm still seeing the
> same problem with the attached bundle.  It keeps loading and reloading
> fields before finally plotting frontogenesis after a considerable wait.
> 
> Jim
> 
> 

Jim,
     I can see the problem when I tried this bundle at home during the weekend. 
Likely, the speed of network in my office covers up the problem.

I will look into it.


Yuan
> 
> On 7/22/15 10:53 AM, Unidata IDV Support wrote:
> >> Yuan:
> >>
> >> These are new bundles.  I've abandoned all the old ones and have
> >> been developing new ones this summer to take advantage of time-matching
> >> and subsetting abilities in the IDV, as well as the better THREDDS
> >> capabilities.  Much of this has gone really well, but I've been
> >> reporting some of the hiccups I've encountered.
> >>
> >> The new bundles are in
> >> http://weather.utah.edu/synoptic/idv/Site_Config/bundles/Real-Time-WX/.
> >> Skip those in the Diagnostics directory as those I haven't updated yet
> >> because of the problem noted below.
> >>
> >> Jim
> > Jim,
> >         I will check in the fixed for the bug and you should be able to get 
> > the nightly earlier this afternoon.
> >         This bug is related to the setting time driver in the animation 
> > widget.
> >         I tried your bundle and the timing seems right to me even though 
> > you will still seeing
> > reading data for many times, those message might not be corrected. I set 
> > the break point for the data
> > reload, and make sure only happen one time at the beginning.
> >
> >
> > Yuan
> >>
> >>
> >> On 7/20/15 4:03 PM, Unidata IDV Support wrote:
> >>>> I'm not sure what you mean by old bundle.  If you mean
> >>>> Frontogenesis-OnlyFr.xidv, yes, that is the one I am working with.
> >>>>
> >>> Jim,
> >>>           I thought you were modifying your old bundles. But, I believe I 
> >>> know what went wrong now.
> >>> The problem is that the IDV don't have the subset information when 
> >>> calling these diagnostic function. Those
> >>> subset information including the time matching, are hiding in its 
> >>> parameters and the parameters can be derived
> >>> itself. So, some of the logic must be wrong in the current release. I 
> >>> will find a solution asap.
> >>>
> >>>
> >>> Yuan
> >>>> We need to find a way to hunt down the source of this problem. Bundles
> >>>> are especially important for more complex, integrated analysis,
> >>>> including diagnostics like frontogenesis.  To have to try and create
> >>>> such an analysis by hand each time would be extremely time consuming.
> >>>>
> >>>> Jim
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> On 7/20/15 3:11 PM, Unidata IDV Support wrote:
> >>>>>> Yuan:
> >>>>>>
> >>>>>> Yes, your bundle works fine.  It loads the fields once and then
> >>>>>> plots it.
> >>>>>>
> >>>>>> However, when I load the bundle that I sent you (attached - note
> >>>>>> that the name OnlyFr means I'm not plotting other additional
> >>>>>> frontogenesis diagnostics, this bundle is meant to plot frontogenesis
> >>>>>> along with several other fields and images), it still is loading t
> >>>>>> fields many times over and over and it takes an inordinate amount of
> >>>>>> time to plot the frontogenesis.  Have you tried this bundle?  Watch how
> >>>>>> many times it loads temperature and u/v.
> >>>>>>
> >>>>>> Jim
> >>>>>>
> >>>>> Jim,
> >>>>>         I did try your bundle and saw the same problem, I don't know 
> >>>>> what causes these multiple reloads. Are you working on the old bundle?
> >>>>>
> >>>>> Yuan
> >>>>>> On 7/20/15 2:52 PM, Unidata IDV Support wrote:
> >>>>>>>>> Hi:
> >>>>>>>>>
> >>>>>>>>> I've been working on some bundles that include frontogenesis
> >>>>>>>>> calculated in IDV using the FRNT function (derived from gempak) 
> >>>>>>>>> using
> >>>>>>>>> potential temperature and wind vectors on an isobaric surface.
> >>>>>>>>>
> >>>>>>>>> When I plot frontogesis directly in IDV, it loads up those fields
> >>>>>>>>> and does the calculations in a reasonable amount of time. However, 
> >>>>>>>>> if I
> >>>>>>>>> then save to a bundle, and reload the bundle, it loads those fields 
> >>>>>>>>> many
> >>>>>>>>> times and takes an excessive amount of time to plot.  The attached
> >>>>>>>>> bundle should illustrate the problem.  It will load in several 
> >>>>>>>>> fields,
> >>>>>>>>> but then hit a wall as it loads and reloads potential temperature 
> >>>>>>>>> and
> >>>>>>>>> wind components over an extended period of time before finally 
> >>>>>>>>> plotting
> >>>>>>>>> frontogenesis.
> >>>>>>>>>
> >>>>>>>>> Jim
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>> Jim,
> >>>>>>>> I tried to reproduce your problem from scratch and no luck. The 
> >>>>>>>> attached bundle uses the time matching and only
> >>>>>>>> calls the frnt function, everything works as expected. Could you 
> >>>>>>>> give it a try ?
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> Yuan
> >>>>>>> Sorry I forgot the bundle.
> >>>>>>>
> >>>>>>> Yuan
> >>>>>>>
> >>>>>>> Ticket Details
> >>>>>>> ===================
> >>>>>>> Ticket ID: NTD-472266
> >>>>>>> Department: Support IDV
> >>>>>>> Priority: Normal
> >>>>>>> Status: Open
> >>>>> Ticket Details
> >>>>> ===================
> >>>>> Ticket ID: NTD-472266
> >>>>> Department: Support IDV
> >>>>> Priority: Normal
> >>>>> Status: Open
> >>>>>
> >>> Ticket Details
> >>> ===================
> >>> Ticket ID: NTD-472266
> >>> Department: Support IDV
> >>> Priority: Normal
> >>> Status: Open
> >>>
> >>
> >
> > Ticket Details
> > ===================
> > Ticket ID: NTD-472266
> > Department: Support IDV
> > Priority: Normal
> > Status: Closed
> >
> 
> 
> 


Ticket Details
===================
Ticket ID: NTD-472266
Department: Support IDV
Priority: Normal
Status: Open