Jeff Whitaker

Not sure what it is about the file, but I can get a segfault trying to read it with [nco](http://nco.sourceforge.net/nco.html) also `ncks -v T_z300m_HHM -d time,11,16 -s '%g ' WN_L2_HHD_v7_UTM_TF1_UTC_2018-01.nc`...

OK, that's good to know. I'd still like to track down why your original code produced an apparently corrupted file. It would be very helpful if you could provide a...

That would be great thanks. Let's leave the ticket open in the meantime.

This and #971 are almost certainly issues with the C library. The python interface for in-memory datasets is just a very thin wrapper on the C library calls.

https://github.com/Unidata/netcdf-c/issues/1489

I have confirmed that the simple python test program in Unidata/netcdf4-python#1018 runs two orders of magnitude slower (using netcdf4-python master and hdf5 1.10.6) when netcdf-c 4.7.4 is linked, as opposed...

simple C program to test (using this [file](https://data.nodc.noaa.gov/thredds/fileServer/ncei/woa/salinity/A5B7/0.25/woa18_A5B7_s16_04.nc)): ```c #include #include #include #define FILE_NAME "woa18_A5B7_s16_04.nc" #define NDIMS 4 #define NLAT 100 #define NLON 100 #define NLVL 102 #define VAR_NAME "s_an"...

Using `keepweakref=True` when opening the Dataset eliminates the segfault for me. ``` import netCDF4 as nc for i in xrange(1, 33): print(i) d = nc.Dataset('issue261.nc',keepweakref=True) ``` This suggests that the...

Of course, using the python context manager will also avoids the segfault (by making sure the file is closed). ``` import netCDF4 as nc for i in xrange(1, 51): print(i)...

The traceback provided by @jdemaria looks similar to one discussed on the h5py list: https://groups.google.com/forum/#!msg/h5py/3v0oBQ3SVkk/qsCwQnfTxuEJ