id,node_id,number,title,user,state,locked,assignee,milestone,comments,created_at,updated_at,closed_at,author_association,active_lock_reason,draft,pull_request,body,reactions,performed_via_github_app,state_reason,repo,type
406971782,MDU6SXNzdWU0MDY5NzE3ODI=,2747,1000x performance regression in transpose from disk between libnetcdf 4.6.1 and 4.6.2,2552981,closed,0,,,3,2019-02-05T20:59:53Z,2020-08-23T18:53:41Z,2020-08-23T18:53:41Z,CONTRIBUTOR,,,,"Having generated `test.nc` as the two-dimensional array from
```python
import numpy as np
import xarray as xr

a = np.random.random((1000, 100))
ds = xr.Dataset({'foo': xr.DataArray(a, [('x', np.arange(1000)), ('y', np.arange(100))])})
ds.to_netcdf('test.nc')
```

I am seeing a huge performance regression from libnetcdf=4.6.1
```python
In [2]: ds = xr.open_dataset('test.nc') 
   ...: %time np.asarray(ds['foo'].transpose('y', 'x')) 
   ...:                                                                                                                
CPU times: user 2.58 ms, sys: 191 µs, total: 2.77 ms
```
in a conda environment created by
```
conda create -n mwe-fast 'libnetcdf=4.6.1' netcdf4 xarray ipython
```
to the environment created by 
```
conda create -n mwe-slow 'libnetcdf=4.6.2' netcdf4 xarray ipython
```
```python
In [5]: ds = xr.open_dataset('test.nc') 
   ...: %time np.asarray(ds['foo'].transpose('y', 'x')) 
   ...:                                                                                                                
CPU times: user 4.09 s, sys: 12 ms, total: 4.11 s
```

Loading into memory mitigates the regression (on `mwe-slow`):
```python
In [2]: ds = xr.open_dataset('test.nc') 
   ...: %time np.asarray(ds['foo'].load().transpose('y', 'x')) 
   ...:                                                                                                                
CPU times: user 1.12 ms, sys: 80 µs, total: 1.2 ms
```

#### Output of ``xr.show_versions()``

<details>

INSTALLED VERSIONS on the working environment `mwe-fast`
------------------

commit: None
python: 3.7.1 | packaged by conda-forge | (default, Nov 13 2018, 18:33:04) 
[GCC 7.3.0]
python-bits: 64
OS: Linux
OS-release: 4.19.0-2-amd64
machine: x86_64
processor: 
byteorder: little
LC_ALL: None
LANG: en_GB.UTF-8
LOCALE: en_GB.UTF-8
libhdf5: 1.10.3
libnetcdf: 4.6.1

xarray: 0.11.3
pandas: 0.24.1
numpy: 1.15.4
scipy: None
netCDF4: 1.4.2
pydap: None
h5netcdf: None
h5py: None
Nio: None
zarr: None
cftime: 1.0.3.4
PseudonetCDF: None
rasterio: None
cfgrib: None
iris: None
bottleneck: None
cyordereddict: None
dask: None
distributed: None
matplotlib: None
cartopy: None
seaborn: None
setuptools: 40.7.3
pip: 19.0.1
conda: None
pytest: None
IPython: 7.2.0
sphinx: None

INSTALLED VERSIONS  on the broken environment `mwe-slow`
------------------
commit: None
python: 3.7.1 | packaged by conda-forge | (default, Nov 13 2018, 18:33:04) 
[GCC 7.3.0]
python-bits: 64
OS: Linux
OS-release: 4.19.0-2-amd64
machine: x86_64
processor: 
byteorder: little
LC_ALL: None
LANG: en_GB.UTF-8
LOCALE: en_GB.UTF-8
libhdf5: 1.10.4
libnetcdf: 4.6.2

xarray: 0.11.3
pandas: 0.24.1
numpy: 1.15.4
scipy: None
netCDF4: 1.4.2
pydap: None
h5netcdf: None
h5py: None
Nio: None
zarr: None
cftime: 1.0.3.4
PseudonetCDF: None
rasterio: None
cfgrib: None
iris: None
bottleneck: None
cyordereddict: None
dask: None
distributed: None
matplotlib: None
cartopy: None
seaborn: None
setuptools: 40.7.3
pip: 19.0.1
conda: None
pytest: None
IPython: 7.2.0
sphinx: None

</details>
","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/2747/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,completed,13221727,issue
464152437,MDExOlB1bGxSZXF1ZXN0Mjk0NDg5OTE4,3079,Fix printing summaries of multiindex coords,2552981,closed,0,,,7,2019-07-04T09:05:23Z,2019-07-11T07:48:56Z,2019-07-10T16:48:01Z,CONTRIBUTOR,,0,pydata/xarray/pulls/3079,"Since merging #2293 summarize_variable displays the first and last few entries,
so we have to pass all along.

MWE:
```python
import xarray as xr
import numpy as np
x = np.arange(0, 90, 0.5)
y = np.arange(0, 90, 0.5)
da = xr.DataArray(np.random.random((len(x), len(y))), [('x', x), ('y', y)])
da.stack(z=('x', 'y'))
```
gives on master:
```
<xarray.DataArray (z: 32400)>
array([0.497724, 0.793955, 0.140193, ..., 0.05734 , 0.884105, 0.479093])
Coordinates:
  * z        (z) MultiIndex
  - x        (z) float64 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ... 0.0 0.0 0.0 0.0 0.0 0.0
  - y        (z) float64 0.0 0.5 1.0 1.5 2.0 2.5 ... 12.5 13.0 13.5 14.0 14.5
```
and on this branch
```
<xarray.DataArray (z: 32400)>
array([0.497724, 0.793955, 0.140193, ..., 0.05734 , 0.884105, 0.479093])
Coordinates:
  * z     (z) MultiIndex
  - x     (z) float64 0.0 0.0 0.0 0.0 0.0 0.0 ... 89.5 89.5 89.5 89.5 89.5 89.5
  - y     (z) float64 0.0 0.5 1.0 1.5 2.0 2.5 ... 87.0 87.5 88.0 88.5 89.0 89.5
```

 - [ ] Closes #xxxx
 - [x] Tests added
 - [ ] Fully documented, including `whats-new.rst` for all changes and `api.rst` for new API
","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/3079/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull