Re: [netcdfgroup] writing/reading a scalar variable in a NETCDF4 type

• To: "Hernan G. Arango" <arango@xxxxxxxxxxxxxxxxxx>
• Subject: Re: [netcdfgroup] writing/reading a scalar variable in a NETCDF4 type
• From: ed@xxxxxxxxxxxxxxxx
• Date: Thu, 29 Jan 2009 08:13:08 -0700 (MST)

"Hernan G. Arango" <arango@xxxxxxxxxxxxxxxxxx> writes:

> I don't understand why the NETCDF-4 selected collective
> operations as the default.  This may be dangerous and may be at
> the root of my problem if collective/independent tests are not done
> everywhere in the low level routines.  MPICH2 sometimes have
> problems (misbehave) during collective I/O operations.
>

Collective is only the default for metadata write operations.

For reading, and for writing data, independent is the default.

Ed

--
Ed Hartnett  -- ed@xxxxxxxxxxxxxxxx