Feature #7843

add support for consuming fixity reports from DuraCloud

Added by Justin Simpson almost 7 years ago. Updated over 5 years ago.

Status:NewStart date:01/20/2015
Priority:HighDue date:
Assignee:-% Done:

0%

Category:-Estimated time:8.00 hours
Target version:-
Sponsored:No Requires documentation:
Pull Request:

Description

The upcoming release of DuraCloud has some new REST api endpoints.

Documentation about the manifest REST API endpoint can be found here: https://wiki.duraspace.org/display/DURACLOUDDOC/DuraCloud+REST+API#DuraCloudRESTAPI-GetManifest

There is a test instance available, credentials in LastPass as DuraCloud QA Server).

The new endpoint we want to use is called GetManifest. This will return either a Tab Separated file, or a BagIt style manifest, listing all the files in a given DuraCloud Space, and their md5sums.

Possible questions to ask DuraCloud include:
can checksums be supplied other than md5 ?
can checksums be supplied for a single file?

History

#1 Updated by Justin Simpson almost 7 years ago

  • Estimated time set to 8.00

#2 Updated by Holly Becker almost 7 years ago

Questions:
  • Can non-md5 checksums be supplied?
    • md5 has decent likelihood of collisions - use sha256 (also what AM uses)
  • Will all checksums be the same format?
  • Is the checksum the same as returned in the Content-MD5 header of HEAD on the file?
  • Can we get checksums for a subset of files? Eg. all files that start with a prefix?
  • How are filenames with a tab in them handled in the TSV?
  • Are checksums generated on demand? Can we see when the checksum was generated? Force a new checksum? Times preferred in ISO 8601
  • Why return the space-id if it's part of the URL? Can multiple space-ids be specified? How?
  • Is this paginated? How many files can be returned at once? How do I get the next page?
  • What encoding are the filenames returned in? UTF8?
    • More generally, how does Duracloud handle encodings of filenames?
  • Suggestion: Use JSON as a format
    • More flexible for future changes

#3 Updated by Justin Simpson over 5 years ago

  • Project changed from Archivematica integration with DuraCloud to Archivematica Storage Service
  • Assignee deleted (Holly Becker)

Also available in: Atom PDF