Dear dcm4che Community,
I am currently planning a migration from dcm4chee v2 to dcm4chee-arc-light (v5) in a hospital environment with a large volume of archived studies spanning multiple years.
We have both systems set up and running locally (Linux environment). My current approach is to migrate data disk-by-disk, based on the entries from the filesystem table in dcm4chee v2.
Before proceeding further, I would appreciate expert input on the following:
Questions:
Is the disk-by-disk + C-MOVE hybrid approach valid and reliable for full-scale migration from v2 to v5?
If not, what strategy do you recommend for ensuring completeness, integrity, and traceability?
Are there known issues when migrating DICOM data from dcm4chee v2 to v5 over the DICOM protocol (C-MOVE) — especially regarding private tags, character sets, or structured reports?
Is there any supported migration tool, script, or documented
method provided by the dcm4che team for database-level migration or
mapping from v2 to v5?
(For example: mapping Patient/Study/Series metadata or preserving StudyInstanceUIDs and AccessionNumbers.)
Do StudyInstanceUIDs need to be remapped or regenerated, or is it best to retain them as-is when moving to v5?
How can I verify that all studies from v2 have been successfully and completely migrated into v5?
Any recommended verification strategies?
What are the recommended practices to ensure both performance and data safety while doing a full migration (thousands of studies, TBs of data)?