File naming convention change

[btupper]

I think a small sample directory with just a few files would be very helpful.  I'm thinking that anything I can do to smooth the transition at my (consumer) end will make life easier.

[OB ODPS - jgwilding]

Does this work?

https://oceandata.sci.gsfc.nasa.gov/opendap/VIIRS/L3SMI/2018/0703/contents.html

The files have the new naming convention but should be considered samples data wise.

john

[btupper]

It’s just right!  Thank you!

[btupper]

Ahoy,

The example was very helpful - thanks again for that.

I'm still trying to noodle out how searches will be performed using THREDDS - and I think I am asking for confirmation of a strategy.

I see that VIIRS is now being output under the new naming/organizing regime.  As shown below, it looks like there is a couple week overlap where both conventions are followed (ala 'jjj' and 'mmdd') before the 'jjj' convention is no longer supported. So all new data will be under the 'mmdd' convention.  My understanding is that as data are reprocessed the archive of old-convention ("classic-convention"?) 'jjj' will be replaced by the new convention - that is that 'jjj' directories and contents will disappear as the comparable 'mmdd' directories and contents appear.

In the meantime, I guess the interim search strategy must first looks for the 'jjj' resources.  If that fails then the search must look for the comparable 'mmdd' resources.  And if that fails then the resource is not available.  Does that sound correct?

Thanks!
Ben

   
    dates      jjj   mmdd
1 2019-09-29 272   NA  
2 2019-09-30 273   NA  
3 2019-10-01 274   1001
4 2019-10-02 275   1002
5 2019-10-03 276   1003
6 2019-10-04 277   1004
7 2019-10-05 278   1005
8 2019-10-06 279   1006
9 2019-10-07 280   1007
10 2019-10-08 281   1008
11 2019-10-09 282   1009
12 2019-10-10 283   1010
13 2019-10-11 284   1011
14 2019-10-12 285   1012
15 2019-10-13 286   1013
16 2019-10-14 287   1014
17 2019-10-15 288   1015
18 2019-10-16 289   1016
19 2019-10-17 290   1017
20 2019-10-18 291   1018
21 2019-10-19 292   1019
22 2019-10-20 293   1020
23 2019-10-21 294   1021
24 2019-10-22 295   1022
25 2019-10-23 296   1023
26 2019-10-24 297   1024
27 2019-10-25 298   1025
28 2019-10-26 299   1026
29 2019-10-27 300   1027
30 2019-10-28 NA    1028
31 2019-10-29 NA    1029

[OB ODPS - towens]

Yes, you are correct.
The jjj directories point to the old files and will be removed shortly.
The Near Real Time files from 2019-09-30 and going forward have been reprocessed with the new naming convention and the V2016.2 configuration and appear in the mmdd directories to provide an overlap with the existing jjj.
Tommy

[btupper]

Hi again,

I'm sorry to be so thick, but I have yet another question.  MODISA SST are now being offered up under the new filename convention (which I like), but they are marked NRT...

https://oceandata.sci.gsfc.nasa.gov/opendap/MODISA/L3SMI/2019/1101/contents.html

Is there document explaining the context of NRT for L3M DAY products?  My understanding of the meaning of NRT is a little wobbly to start with, and perusing the forum/docs (https://oceancolor.gsfc.nasa.gov/docs/format/Ocean_Level-3_SMI_Products.pdf) hasn't switch the light bulb on. 

Thanks,
Ben

[OB.DAAC - SeanBailey]

Ben,

The NRT label is to make explicit what was buried in the file metadata before the name change.
"NRT" is what we often refer to as "QuickLook" (see https://oceancolor.gsfc.nasa.gov/forum/oceancolor/topic_show.pl?pid=34121)

These files *will* be regenerated once we have all the necessary bits to remove the NRT moniker.

Sean

[btupper]

I see.  So ...

AQUA_MODIS.20191101.L3m.DAY.NSST.sst.4km.NRT.nc

will become

AQUA_MODIS.20191101.L3m.DAY.NSST.sst.4km.ThisOneIsWayBetter.nc

or more likely

AQUA_MODIS.20191101.L3m.DAY.NSST.sst.4km.nc

Thanks!
Ben

[OB.DAAC - SeanBailey]

Yes, the later - but I do kinda like the look of the former :grin:

Sean