You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When any new FIleCollection is being registered, the size and num_files attributes are set to nil for a short period of time until, in the background, the system has finished counting the files and the sum of their sizes.
So there is a short window when a user can access the 'show' page of such a FileCollection, and because the size i not set, they can click the 'download' button and trigger a background synchronization and TAR making (what is commonly done during a dowload) even for very very large collections.
The download action should refuse to trigger a download on any file that doesn't yet have a size set (and warn the user about this).
The text was updated successfully, but these errors were encountered:
Some context. A user registered a huge 40+ gig FileCollection with 280,000 files on a personal data provider. The user went to the show page while the sizes and number of files were still being counted and clicked the 'download' button. CBRAIN proceeded over the next several hours to sync the data set to cache and create a huge TAR file that was not sent anywhere in the end, because the user's browser had disconnected anyway.
When any new FIleCollection is being registered, the
size
andnum_files
attributes are set to nil for a short period of time until, in the background, the system has finished counting the files and the sum of their sizes.So there is a short window when a user can access the 'show' page of such a FileCollection, and because the size i not set, they can click the 'download' button and trigger a background synchronization and TAR making (what is commonly done during a dowload) even for very very large collections.
The download action should refuse to trigger a download on any file that doesn't yet have a size set (and warn the user about this).
The text was updated successfully, but these errors were encountered: