Why does the .zattrs file disappear after rechunk?

[Peterpan-Coder]

Hi, I'm learning to use Zarr, but I've had some problems and I can't find the answer on the official website. I want to get an answer.
I have a climate multidimensional data in NetCDF format. I converted it to Zarr format using XArray. Then I tried using Zarr's rechunk feature.

I rechunk the air array using the following code.

def print_nc_info(ncpath):
    ds = xr.open_dataset(ncpath)
    compressor = zarr.Blosc(cname='zstd', clevel=3)
    encoding = {vname: {'compressor': compressor} for vname in ds.data_vars}
    ds.to_zarr(store='zarr_example', encoding=encoding, consolidated=True)  
    print(xr.open_zarr('zarr_example'))
    zr = zarr.open('zarr_example')
    air = zr['air']
    tmp = air[...]
    zarr.array(tmp, chunks=(6, 17, 73, 72), store='zarr_example/air', overwrite=True)
    print(xr.open_zarr('zarr_example'))
    
if __name__ == '__main__':
    print_nc_info('D:\\test\\air.mon.1981-2010.ltm.nc')

The result is as follows. In the new air folder. The .zattrs file disappeared and I couldn't open the data again with xarray.

Re-chunk

Re-chunk is to improve transfer efficiency, but I now have problems with the file after recunk. I want to understand zarr‘ s rechunk feature better

[Peterpan-Coder]

I tried using Rechunker and the problem was solved.