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

[netCDF #YEI-230852]: adding attributes after data with nc_redef

If the file is in netcdf-4 format, then the performance consequences
should be very small -- only the cost of writing the new attributes.

If the file is in netcdf-3 format, then
basically what is going to happen is that the existing file data
will be moved in the file to make room for the new metadata
(i.e. the new attributes in this case). So the cost will be proportional
to the amount of data. Also, the new metadata will need to be re-written
to the file, and the cost to do that will depend on the amount of
metadata.
At a guess, the total cost will be about the same as
reading and writing the file once.
There is an exception to this. If you are not writing new attributes
and the new values you give to existing attributes is no larget
than the original size, then the only cost is to write the attribute.
The metadata and data will not be re-written.

> 
> On your website one can read with regard to "nc_enddef" the following:
> "This call may involve copying data under some circumstances." Could you
> please tell me what performance implications it may have and under which
> circumstances? Or more specifically, could you please tell whether there
> are any performance or otherwise possible complications with re-entering
> the define mode and adding attributes after the file has been already
> populated with (big) data?
> 
> Thanks,
> 
> Alex
> 
> 
> 

=Dennis Heimbigner
  Unidata


Ticket Details
===================
Ticket ID: YEI-230852
Department: Support netCDF
Priority: Normal
Status: Open
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.