Hi Yuan and John,
For some reason I have the call to GridDataset.sync() commented out in
RAMADDA. I vaguely recall problems with it a while back.
John - in the latest version of the CDM does sync() reread the file and
handle when there is a new file?
Thanks,
Jeff
On Wed, Oct 30, 2013 at 4:46 PM, Yuan Ho <address@hidden
<mailto:address@hidden>> wrote:
Jeff and Don,
We verified the same problem in the motherlode, and remember
Tom's African project has the same issue. Is there a possible other
than reboot the server.
By the way, the thumbnail image of ramadda catalog in not
working in the IDV, any idea?
Yuan
-------- Original Message --------
Subject: New Ticket - [IDV !HAU-717700]: IDV v4.1 and RAMADDA:
Loading new data with same file name as old data plots the old data
Date: Wed, 30 Oct 2013 15:22:29 -0600
From: David P Dempsey <address@hidden
<mailto:address@hidden>>
Reply-To: address@hidden
<mailto:address@hidden>
To: address@hidden <mailto:address@hidden>
New Ticket: IDV v4.1 and RAMADDA: Loading new data with same file
name as old data plots the old data
Yuan, Julian, etc.,
I'm using IDV v4.1 on an iMac running Mac OS X v.10.8.5.
I'm loading and plotting geographic data, including topographic
elevations, for a domain generated locally by the WRF Preprocessing
System (WPS). I'm loading the data via one of our local RAMADDA
servers (http://virga.sfsu.edu:8080/__repository/thredds
<http://virga.sfsu.edu:8080/repository/thredds>). The file is a
NetCDF file (geo_em.CenCal_BayArea.nc
<http://geo_em.CenCal_BayArea.nc><http__://virga.sfsu.edu/data/wrf/__geo_em.CenCal_BayArea.nc
<http://virga.sfsu.edu/data/wrf/geo_em.CenCal_BayArea.nc>>). I can
load and plot this just fine (a bundle is attached:
Dempsey_SFBayAreaTopography___RAMADDA.xidv.)
However, a problem arises when I create a new geographic data set
with the same size and containing the same types of fields but
shifted in location, and overwrite the previous NetCDF file with the
new data, keeping the same file name. To plot this geographically
shifted data set in the IDV after plotting the original one, I first
remove the original displays and loaded data by selecting Edit >
Remove All Displays and Data. Then I access and load the new file
(which has the same name as the old, overwritten file), and plot the
result using exactly the same plotting specifications as the
original one.
Unfortunately, the plot that I get is identical to the original one.
This is true even though the contents of the files are not the same
(I checked to make sure)--the location metadata in each of the two
files are exactly what I want them to be (not the same). Hence, the
IDV still appears to be plotting the original data, not the new
data, even though I've presumably removed the original data and
loaded the new data.
I get the same results if I first remove the original data source by
<Command>-clicking on the data source in the Dashboard > Field
Selector > Data Sources, and selecting "Remove Data Source". Ditto
if I remove the data from the IDV by either means, then quit the
IDV, restart, and try again. Ditto if I refresh the listing of files
offered by RAMADDA after overwriting the original file but before
selecting and loading the new version. Ditto if I don't remove the
display of the original plot, or don't remove either the display or
the data source before loading the replacement data and plotting it.
On the other hand, if I give the new file a different name from the
original one, load it, and plot it, I get a plot of the shifted
domain, just as I expect.
On the third hand, if I access and load the new data using HTTP or
as a local file instead of via RAMADDA, I get a plot of the shifted
domain, as desired.
On the fourth hand, if I put the two different data sets in files
with the same name but in two different directories (e.g., SFSU
Weather Data > wrf and in SFSU Weather Data > wrf > temp_demo), and
plot the second one after removing the data and displays created
from plotting the first one, then the second plot is successful.
And finally, if I access the original file via HTTP or as a local
file, then remove the data and display from the IDV, then overwrite
the original data with the new data (keeping the file name the
same), and finally access and load the new file via RAMADDA, the new
plot is correct.
To summarize: Using RAMADDA, there is a problem when loading and
plotting two different sets of data in succession stored files that
have the same file name and reside in the same directory. The second
plot is identical to the first plot, even when the data and display
from the first plot are removed before loading the replacement data.
A final note: when I refresh the listing of the directory containing
the data in RAMADDA (SFSU Weather Data > wrf), RAMADDA jumps back to
a parent directory (containing SFSU Weather Data) two levels higher
up, though when I navigate back down to the directory I want (wrf),
the listing is indeed refreshed. This is annoying behavior, though!
-- Dave
P.S. Yuan, you're probably fully engaged responding the significant
feedback being provided about the new image plotting enhancements,
to this probably feels like a distraction! (I wonder if Jeff
McWhirter is ultimately going to have to deal with it, if it turns
out to be primarily a RAMADDA problem.)
******************************__******************************__***
* Dr. Dave Dempsey | ^ ___ \|/
*
* Dept. of Earth & Climate Sciences | ) ^ /||_||\ --0-- *
* San Francisco State University | ) ) / ||_|| \
/|\ *
* 1600 Holloway Ave. | ) ) / ||_||
\ *
* San Francisco, CA 94132 | ) ) / ||_||
\ ^ *
* |
) ) ) ||_|| \ *
* Phone: (415) 338-7716 <tel:%28415%29%20338-7716>
| ) ) )~||~||~~~~ \~~*
* FAX: (415) 338-7705 <tel:%28415%29%20338-7705>
| ) ) ) ) ~ ~ ~ ~ ~ ~*
* Email: address@hidden
<mailto:address@hidden><mailto:address@hidden
<mailto:address@hidden>> | ) ) ) ) ) ~ ~ ~ ~ *
******************************__******************************__***
Ticket Details
===================
Ticket ID: HAU-717700
Department: Support IDV
Priority: Normal
Status: Open
Link:
https://www.unidata.ucar.edu/__esupport/staff/index.php?_m=__tickets&_a=viewticket&__ticketid=22912
<https://www.unidata.ucar.edu/esupport/staff/index.php?_m=tickets&_a=viewticket&ticketid=22912>