html_url,issue_url,id,node_id,user,created_at,updated_at,author_association,body,reactions,performed_via_github_app,issue
https://github.com/pydata/xarray/issues/5902#issuecomment-953351408,https://api.github.com/repos/pydata/xarray/issues/5902,953351408,IC_kwDOAMm_X8440vzw,2448579,2021-10-27T22:16:17Z,2021-10-27T22:18:33Z,MEMBER,"(warning: untested code)

Instead of looking at all of `self.variables` we could
``` python
nonindexes = set(self.variables) - set(self.indexes)
# or alternatively make a list of multiindex variables names and exclude those

# then the condition becomes
any(is_duck_dask_array(self.variables[v].data) for v in nonindexes)
``` ","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1037894157
https://github.com/pydata/xarray/issues/5902#issuecomment-953352129,https://api.github.com/repos/pydata/xarray/issues/5902,953352129,IC_kwDOAMm_X8440v_B,2448579,2021-10-27T22:17:39Z,2021-10-27T22:17:39Z,MEMBER,PS: It doesn't seem like the bottleneck in your case but #5582 has an alternative proposal for unstacking dask arrays.,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1037894157