* Repository meets sustainability expectations.
* What data will be exposed?
* Content mutability.
* Software needs:
* DataONE software stack or implement new web service in existing applications (custom stack)?
* Full repository or Slender Node?
* metadata only or science data + replication?
* Repository's in-house preferences? (Java? Python?)
* Consistency of metadata quality.
* Level of functionality (tier).
* Confirm chosen deployment path.
* Communicate knowledge of technical requirements for selected deployment. These can include:
* SSL certificate.
* Apache web server with Python 2.7 or Java (depending on stack).
* Confirm metadata format. Communicate expectations for metadata quality.
* Specify in advance whether a separate test installation will be maintained in parallel with production. This is important for planning the deployment. A separate test will need distinct subdomain and SSL cert if not using wildcard.
* Receive and validate sample metadata records from member node.