Datasets with same name

[ardhimanshu]

Hi @adrianq , Using our earlier DC app version (not the 2.31 one) one of our users Destelia tried to edit an existing dataset (IQFM1950...). She got Error 500 (Internal server error) while saving. So she tried 3 more times. Ultimately, she ended up having 4 copies of the dataset with same name. I noticed that uids are different though. The questions / issues are:

1. The error 500: I believe it is not to do with DC app but would be great to know what causes such error so we can look into relevant things on our side and correct it.
2. Having datasets with same name... Can you please look into it (screenshots attached).. Thanks and regards, Himanshu

[ardhimanshu]

[tokland]

Hi @ardhimanshu,

Firstly, I tried to reproduce the problem, so I created a user TestEst1, a clone of user Destelia. Then I went to the edit of the data set:

https://gors.nrc.no/api/apps/dataset-configuration%20(3)/index.html#/datasets/edit/UYV8UDUc7N3?_k=og3iax

I did not change anything. On the last step, I clicked SAVE, and it was saved fine: Dataset saved successfully.

Did you make any change to the user (groups, roles, authorities) or the data set itself?

If the user made some changes to the data set on Edit before clicking save, we would need those changes.

Feel free to delete those duplicated data sets.

[ardhimanshu]

Hi @tokland, We are waiting for the user response. Will get back on that. However, the key challenge still is:

1. Saving datasets with same name so the user can become confused. As you can see there are 4 versions of this dataset (though with different uid).
2. Similar saving error occured for the same user (Destelia) again for IQFM1952 (OCHA) Dataset leading to creating two datasets with same name.

[tokland]

Hi @ardhimanshu,

Saving datasets with same name so the user can become confused.

Absolutely, but the thing is that an edit->save operation should never create a new data set, as it POST the existing one with its id. We need to understand the context of this edition to reproduce it and find the source of the problem.

Saving datasets with same name so the user can become confused.

I made the same test: I edited the original dataset (snpTRqKn0rV), changed nothing, saved, and no new data set is created. So there must be something different on how this user does it.

If the user finally reports that she is doing exactly this, no changes and edit, then we should test with her user and not our cloned one, but for now I am assuming they behave the same.