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/3306#issuecomment-531499393,https://api.github.com/repos/pydata/xarray/issues/3306,531499393,MDEyOklzc3VlQ29tbWVudDUzMTQ5OTM5Mw==,1197350,2019-09-14T17:47:10Z,2019-09-14T17:47:10Z,MEMBER,"What if you just use a dask local cluster, rather than a distributed cluster? Then you can just write to a local directory.

And what if you don’t use a distributed cluster at all, just the threaded scheduler?

In my experience with these problems, by systematically removing layers of complexity from the scenario, we often come to the root of the issue 

> On Sep 14, 2019, at 11:03 AM, Aimee Barciauskas <notifications@github.com> wrote:
> 
> @rabernat good points. One thing I'm not sure of how to make reproducible is calling a remote file store, since I think it usually requires calling to a write-protected cloud storage provider. Any tips on this?
> 
> I have what should be an otherwise working example here: https://gist.github.com/abarciauskas-bgse/d0aac2ae9bf0b06f52a577d0a6251b2d - let me know if this is an ok format to share for reproducing the issue.
> 
> —
> You are receiving this because you were mentioned.
> Reply to this email directly, view it on GitHub, or mute the thread.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,493058488
https://github.com/pydata/xarray/issues/3306#issuecomment-531489772,https://api.github.com/repos/pydata/xarray/issues/3306,531489772,MDEyOklzc3VlQ29tbWVudDUzMTQ4OTc3Mg==,1197350,2019-09-14T15:44:32Z,2019-09-14T15:44:32Z,MEMBER,"Does the problem only arise when writing to s3fs? Or can you reproduce it writing to a local Zarr directory store?

Sent from my iPhone

> On Sep 14, 2019, at 11:03 AM, Aimee Barciauskas <notifications@github.com> wrote:
> 
> @rabernat good points. One thing I'm not sure of how to make reproducible is calling a remote file store, since I think it usually requires calling to a write-protected cloud storage provider. Any tips on this?
> 
> I have what should be an otherwise working example here: https://gist.github.com/abarciauskas-bgse/d0aac2ae9bf0b06f52a577d0a6251b2d - let me know if this is an ok format to share for reproducing the issue.
> 
> —
> You are receiving this because you were mentioned.
> Reply to this email directly, view it on GitHub, or mute the thread.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,493058488
https://github.com/pydata/xarray/issues/3306#issuecomment-531437477,https://api.github.com/repos/pydata/xarray/issues/3306,531437477,MDEyOklzc3VlQ29tbWVudDUzMTQzNzQ3Nw==,1197350,2019-09-14T02:00:54Z,2019-09-14T02:00:54Z,MEMBER,"@aidanheerdegen - thanks so much for posting this issue! I think a lot of people run into these sorts of problems, so it's useful to have an example on the issue tracker.

These problems can unfortunately be very hard to debug. If other developers can quickly reproduce your exact same error on their own systems, then we can try to dig deeper. However, I can't run the code you shared. If I paste it into a notebook, I get
```
FileNotFoundError: [Errno 2] No such file or directory: b'/home/jovyan/cmip6-bot/2018/153/20180602090000-JPL-L4_GHRSST-SSTfnd-MUR-GLOB-v02.0-fv04.1.nc'
```
Your example requires your files, which I don't have. I noticed some later examples point to a podaac opendap sever, commented out with the comment
```
# This works fine
# base_url = 'https://podaac-opendap.jpl.nasa.gov:443/opendap/allData/ghrsst/data/GDS2/L4/GLOB/JPL/MUR/v4.1/'
```
but I couldn't tell if I was supposed to run that part or not.

So unfortunately I have to ask you for some tweaks to your question.
Could you either:
1. edit this example so that it points to an opendap server or other globally accessible endpoint on the internet? OR
1. replace the real data with synthetically generated data (e.g. use `dask.random.random` to create arrays instead of loading from disk)

Although 2 is a pain, it actually usually helps surface bugs by removing part of the I/O.

Thanks again for your contribution.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,493058488