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

Re: 20031118: netcdf data problem (fwd)

------- Forwarded Message

Date:    Thu, 12 Feb 2004 14:37:36 +0100
From:    "Birgit Heese" <address@hidden>
To:      "Russ Rew" <address@hidden>
cc:      <address@hidden>
Subject: Re: 20031118: netcdf data problem (fwd) 

Hei Russ,

it's working! Thank you very much!

Good luck! And please keep me informed!

Best wishes, Birgit

----- Original Message ----- 
From: "Russ Rew" <address@hidden>
To: "Birgit Heese" <address@hidden>
Sent: Thursday, February 12, 2004 1:57 PM
Subject: Re: 20031118: netcdf data problem (fwd)


> >To: <address@hidden>
> >From: "Birgit Heese" <address@hidden>
> >Subject: netcdf data problem
> >Organization: MIM
> >Keywords: 200311171529.hAHFT5Ob010057
>
> Birgit,
>
> > thank you for the suggestion using the nc_share flag.
> > But I am using Fortran. My open command is:
> >
> > nf_open("file.dat", nf_write, ncid)
> >
> > and nc_share is not known then.
> > Is there a coressponding flag  like "nf_share"?
> > I tried nf_share, but it didn't work.
>
> Yes, sorry I didn't notice you were using the Fortran interface.  For
> the Fortran-77 interface, the flag is "NF_SHARE" and it's documented
> as one of the flags used in the NF_OPEN call:
>
>
http://www.unidata.ucar.edu/staff/russ/netcdf/guidef/guidef-10.html#HEADING10-112
>
> and would be used like this:
>
>   STATUS = NF_OPEN('file.nc', IOR(NF_WRITE, NF_SHARE), NCID)
>
> If you are using the Fortran-90 interface instead, the "NF90_SHARE"
> has a similar purpose and is documented as one of the flags used in
> the NF90_OPEN call:
>
>
http://www.unidata.ucar.edu/packages/netcdf/f90/Documentation/f90-html-docs/guide5.html#1034444
>
>   STATUS = NF90_OPEN(PATH = "file.nc", CMODE = IOR(NF90_NOWRITE,
NF90_SHARE), NCID = NCID)
>
> Even though the share flag is intended to support concurrent access,
> it also is currently the best workaround I know of for this bug, since
> it turns off caching and buffering which is where the bug occurs.  I
> still hope to get a fix in for this when we release version 3.6.0.
>
> --Russ
>
>
>

------- End of Forwarded Message