-
Task
-
Resolution: Unresolved
-
Minor
-
None
-
None
-
None
-
Sprint 293: Deluxe Data
Re-registering the same file (e.g. manual import with HTTP) creates a conflicting file rather than updates the original one. According to the docs: The registration of the same file can be repeated, which updates the previous metadata with the new information. In most cases, the metadata is overwritten with new values, with exception of extended attributes - previous values are merged with new ones.
TODOs:
- check behaviour on different storage backends
- auto scan must be okay because we use it all the time and there are no conflicts?
- add tests for this behaviour and regression
- while we're at it, add tests checking the safeguards in this PR:
- https://git.onedata.org/projects/VFS/repos/op-worker/pull-requests/4185/diff#src%2Fmiddleware%2Fplugins%2Ffile%2Fhandlers%2Ffile_middleware_create_handler.erl
- https://git.onedata.org/projects/VFS/repos/op-worker/pull-requests/4185/diff#src%2Fmodules%2Ffile_registration%2Ffile_registration.erl