Re: Crash while writing to netcdf4

From: Dave Allured - NOAA Affiliate <dave.allured_at_nyahnyahspammersnyahnyah>
Date: Thu Mar 20 2014 - 11:36:33 MDT

I forgot to mention this. After each crash, the target Netcdf4 file is
broken and can no longer be read or written. This makes for a tedious
debug cycle because the whole original file must be recopied.

--Dave

On Thu, Mar 20, 2014 at 11:18 AM, Dave Allured - NOAA Affiliate <
dave.allured@noaa.gov> wrote:

> NCL team,
>
> I have an NCL program that consistently crashes with segmentation fault
> when writing to certain large Netcdf4-classic files. This program worked
> successfully for dozens of files over the past year, so I am fairly
> confident in the NCL code itself.
>
> In detail, the crashes occur when I try to append records to existing
> files. The files contain a single large 4-D data variable, several smaller
> ones, and an unlimited time dimension. Netcdf4 chunking and compression
> are used. In one case, the file size is 89 Gb, and the seg fault occurs
> when I try to quit the program, after all update data seems to be
> successfully appended. I got the same behavior on two different Linux
> machines and one Mac. More details on request.
>
> As a shortcut to debugging, may I please request an interim binary copy of
> NCL? I am using the current release, 6.1.2, which is over a year old. I
> understand that there have been several bug fixes to the Netcdf and HDF5
> libraries, so I would like to see if some of these might fix the problem.
> I need either a 64-bit Linux version, or a 64-bit Mac version, your
> choice. Thank you.
>
> --Dave
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Thu Mar 20 11:36:46 2014

This archive was generated by hypermail 2.1.8 : Fri Mar 21 2014 - 15:49:21 MDT