Personal accounts are set up similarly with export DATAHUB_JSON, the content of the configuration file being populated when running data login and logging in with Google.
data push succeeds and a new revision appears on DataHub (but fails to validate). Here's the output of data push --debug for a dataset that was successfully pushed before (e.g. see the dataset here https://datahub.io/core/unicode-emojis):
Processing this revision has failed. See below for details:
Copying source data
assembler.load_private_resource
Traceback (most recent call last):
File "/usr/lib/python3.6/site-packages/datapackage_pipelines_assembler/processors/load_private_resource.py", line 35, in <module>
PrivateResourceLoader()()
File "/usr/lib/python3.6/site-packages/datapackage_pipelines_assembler/processors/load_private_resource.py", line 19, in __init__
if PKGSTORE_BUCKET in url:
TypeError: argument of type 'NoneType' is not iterable
This is not a one-time occurrence as can be seen here:
assembler.dump_to_s3
Traceback (most recent call last):
File "/usr/lib/python3.6/site-packages/datapackage_pipelines_assembler/processors/dump_to_s3.py", line 32, in <module>
MyS3Dumper()()
File "/usr/lib/python3.6/site-packages/datapackage_pipelines/lib/dump/dumper_base.py", line 47, in __call__
finalizer=self.finalize
File "/usr/lib/python3.6/site-packages/datapackage_pipelines/wrapper/wrapper.py", line 64, in spew
for res in resources_iterator:
File "/usr/lib/python3.6/site-packages/datapackage_pipelines/lib/dump/dumper_base.py", line 150, in handle_resources
self.handle_datapackage(datapackage, parameters, stats)
File "/usr/lib/python3.6/site-packages/datapackage_pipelines/lib/dump/dumper_base.py", line 212, in handle_datapackage
location = self.write_file_to_output(temp_file_name, 'datapackage.json')
File "/usr/lib/python3.6/site-packages/datapackage_pipelines_aws/s3_dumper.py", line 46, in write_file_to_output
Key=key)
File "/usr/lib/python3.6/site-packages/datapackage_pipelines_assembler/processors/dump_to_s3.py", line 27, in put_object
datetime.datetime.now()
File "/usr/lib/python3.6/site-packages/filemanager/models.py", line 103, in add_file
assert sf.owner == owner
AssertionError
When I use
data-cli
, I want to be able todata push
with that tool successfully so I can publish my dataset to DataHub.DataHub shows the message
Processing this revision has failed.
when pushing a dataset. See Analysis section below for more on this problem.Acceptance
Tasks
Analysis
Organization accounts are set up locally like so before attempting to push:
Example content for
~/.config/datahub/config.json
:Personal accounts are set up similarly with
export DATAHUB_JSON
, the content of the configuration file being populated when runningdata login
and logging in withGoogle
.data push
succeeds and a new revision appears on DataHub (but fails to validate). Here's the output ofdata push --debug
for a dataset that was successfully pushed before (e.g. see the dataset here https://datahub.io/core/unicode-emojis):Opening the resources, e.g. https://s3.amazonaws.com/rawstore.datahub.io/d20273693fdc7ad3d3f061f3eb8045ff.json shows they are successfully uploaded. However, DataHub shows this:
https://datahub.io/sebastien.lavoie/unicode-emojis/v/7 (the code from this error can be found here https://github.com/datopian/assembler/blob/master/datapackage_pipelines_assembler/processors/load_private_resource.py#L19)
This is not a one-time occurrence as can be seen here:
For some unknown reason, https://datahub.io/sebastien.lavoie/unicode-emojis/v/5 worked while
/v/3
,/v/4
and/v/6
didn't.Also, https://datahub.io/core/unicode-emojis/v/194 presented an additional error (appearing twice in the output on DataHub), for which the code can be found here https://github.com/datopian/assembler/blob/master/datapackage_pipelines_assembler/processors/dump_to_s3.py#L27 :
After having failed for about 2 hours straight,
data push
was successful here for a personal account https://datahub.io/sebastien.lavoie/unicode-emojis/v/9 as well as here https://datahub.io/core/unicode-emojis for an organization account.At this moment, the issue seems to be no longer reproducible.