,MetaArchive,DuraCloud,Preservica,DPN,APTrust,Chronopolis,Arkivum,Texas Digital Library,Ex Libris Rosetta
URL,http://metaarchive.org/,http://duracloud.org/,http://preservica.com/,http://dpn.org/,http://aptrust.org/,https://library.ucsd.edu/chronopolis/infrastructure/index.html,http://arkivum.com/,,
Functionality,,,,,,,,,
Setup and Configuration,,,,,,,,,
Installation,Members hosting a server storage node are provided instructions and support to setup server according to technical specifications to connect to network.,Customers / users setup account and can then utilize web dashboard to manually add / ingest content or download and install DuraCloud Sync tool to automatically ingest content.,"Hosted ""Cloud Edition"" service requires no install for main software (see ingest for details on other local software options). Enterprise Edition requires local server and storage (eg.Windows / Linux servers, Network Attached Storage, etc) and default MySQL database setup for metadata. For on premise version, can choose from a number of other databases as well: postgres, oracle, etc. ",,,No installation needed - just needs to be accessible via ssh or via DuraCloud,"No installation needed, bu arkivum sends hardware (server) for customer to install",Yes. duracloud.,"Yes, do installation for the customer. "
,,,,,,,,,
Content,,,,,,,,,
Supports all file formats and/or content types,Yes,"Yes, all file formats and content types are supported","Yes, but users can customize file format requirements and policies",Yes,Yes,Yes,Yes,Yes,Yes.
Supports all metadata schemas and structures,Yes,"Yes, No specific requirements for metadata schemas and/or structures",Yes,Yes,Yes,Yes - utlizes Bagit and recommends supplementary collection data to be included.,Yes,Yes,"Dublin Core is primary metadata schema. Others can be added as ""source"" metadata. Logical and physical structmaps can be added to the content. "
Supports large files (e.g. audio / video),"Yes, but files may be broken into smaller chunks ","Yes, files that are large will be transferred as chunks. No limit to file size. ",Yes,Yes,Yes,Yes - currently working with large data archives,Yes,Yes--maybe 5gb.,Yes.
,,,,,,,,,
Ingest,,,,,,,,,
Ingest process ,"1. Public live web-based content is crawled and ingested via LOCKSS plugins that are setup with specific parameters to automatically ingest content from repository websites (e.g. CONTENTdm, DSpace, Digital Commons)   2. Non-public content is made available to network on web server in simple directory listing (as Bagit bags if preferred)",Content can be added manually via drag and drop to web dashboard or automatically via the DuraCloud Sync Tool  https://wiki.duraspace.org/display/DURACLOUDDOC/Getting+Started+with+DuraCloud,Users can: 1. Upload content directly via web dashboard 2. via locally installed SIP Creator 3. via locally or network installed Transfer Agent 4. via CE Bulk Upload Service,DuraCloud software used to upload content,"APTRust uses the bagit spec. Partners bag their own content and submit their bags through the custom, APTrust API. They can track the bag’s progress up through final ingest into AWS.",DuraCloud software used to upload content. Upload only via the UCSD node.,Customer moves files onto network file share. Perpetua/Archivematica users gor throgh AWS,GUI interface for uploading 1 item at a time trhough duracloud installation.  OR also desktop client that syncs to system from a folder. OR command line that pushes from server to duracloud,"Support manual ingest with web-based interface and semi-automatic ingest using csv, xml, oai-pmh harvester and bagit. Provides also deposit APIs. Once content is in the system it goes through a set of processes including decomposition, virus scanning, fixity checks, format identification, technical metadata extraction, risk management, enrichment processes, and curation and assessment processes. "
Supports ingest of Bagit bags,"Yes, via LOCKSS plugin designed to ingest Bagit bags setup with a specific MetaArchive profile","Yes, but no specific info about Bag requirements. See them as just another file. ",Yes,Specific requirements for configuring Bagit bags for ingest https://docs.google.com/document/d/1JqKMFn9KfeIMAAEdOGQr6LZPqNWx8Qubi12uoUXi2QU/edit,Yes,Yes - specifics available to users,Yes. Suported by Archimetica by perpeuta  and so does the archviing service,"Yes, but only for DPN ingest",Yes.
,,,,,,,,,
Storage,,,,,,,,,
Multiple copies,Seven copies for all membership levels,Most subscription plans include two copies of content stored on two different cloud storage providers. Subscriptions plans for DPN and/or Chronopolis would include 3-4 copies total.  https://wiki.duraspace.org/display/DURACLOUDDOC/DuraCloud+Storage,"Cloud edition includes storage of one copy in Amazon S3 or Amazon Glacier. Enterprise Edition included options for local storage, and connector to Amazon S3, Glacier or Microsoft Azure ",Three copies,6 total ,Three copies,Three copies (Two UK stored data copies accessible online; One UK stored copy held offline with a thrid party ESCROW data holding company),"Recommend use of glacier and S3, so 2 technologies and 2 locations. If you do both, you should have 6 copies, but it is abstracted so they cannot see it. If they are using DPN it is different.",Responsiblity of customer. Can use whatever storage they like and define the required copies. 
Geographic distribution of copies,"Seven copies are distributed among multiple membership locations, including internationally","Most subscriptions plans include at least two copies that are geographically distributed, but details would depend on cloud storage provider locations.","Minimal. Cloud edition with Amazon S3 has option for copies distributed up to 10km apart. Enterprise Edition has more options for distribution by using multiple cloud storage services. No limit and can use any storage option: tapes, AWS, Oracle, on premise, etc. ",Distribution via replication nodes ,"3 West coast, 3 East coast in a combination of S3 and Glacier.",Distributed geographically among partner sites. ,All stored in UK. Distributed in that one copy is held with third party company,west coast glacier and east coast S3.,Customer's responsibility.
Regular fixity checks and monitoring,"Yes, LOCKSS software includes regular automatic fixity checks and verification via voting and polling procedures","Yes, but details including frequency and mechanisms unclear https://wiki.duraspace.org/display/DURACLOUDDOC/DuraCloud+Features","Yes, files checked on an ongoing basis.",,Roughly every 90 days.,Yes - all data in the system undergo ingegrity monitoring using the Audit Control Environment (ACE),"Yes, monthly ""checks and maintenance updates"" and ""annual data retrieval and integrity checks""",currenltly-- every 6 months on S3,Can be run by Rosetta or by the storage system and Rosetta will receive the results. 
Hardware / infrastructure,"All members (except for Preservation members who choose option to pay technology fee) purchase and host server storage node infrastructure to run LOCKSS software and replicate other member's content. The implementation of this server is flexible, but must meet current technical specifications. Additionally, a web server (local or cloud-based) may be needed to stage content and make available for ingest into the storage network ",Hosted service based on multiple cloud storage service providers. Run primarily an AWS but also have storage options from other vendors. ,Cloud edition includes hosted infrastructure. Enterprise utilizes local infrastructure with connectors to cloud storage services. There are no specific requirements for choosing cloud storage over another option for enterprise version. ,,"S3 is AWS’s spinning disk, Glacier is not known. Used to be tape but it is believed to be something else now. Important factor is that it is different than the S3 stack.",Hosted infrastructure via DuraCloud. Each location stores data on spinning disk.,"""local"" cloud hosting or managed service option","hosted via AWS, ",Any storage can be used by and connected to Rosetta using the Storage Abstraction Layer.
,,,,,,,,,
Security,,,,,,,,,ISO 27001 certified
,All member server caches communications are SSL encryption. Server caches setup include firewall and port settings to restrict access only to other servers in the network ,"Content security based on cloud storage service provider protocols (e.g. Amazon, RackSpace). All content required to flow via https so it is all encrypted. Application also requires login and access control settings. ",Details unclear beyond access roles and rights. All users must authenticate with user name/password. User account is set up with series of roles which is equated to security tags. Users can only see what they are allowed to based on their roles and security tags and can be as granular as file level permissions.,,"AWS provides data integrity and siloing so that data can be secure. This follows the US security protocols established in the last administration for AWS (data. gov ) We will be doing a security audit in the coming year. 
",Dark archive that is password protected. Data at each location is limited to system administrators. Nodes are in data centers each with security protocols in place.,File encryption with customer supplied encryption keys,"that which is inherent to Amazon, SSH. come back to this via email",
,,,,,,,,,
Access,,,,,,,,,
Restricted authenticated access ,Only designated system administrators have user accounts for member server caches. No member is permitted to access another member’s stored collections. Login pages are only accessible to the host Member institution and the MetaArchive central staff. SSH is required for remote access. ,Access control for users is managed via the DuraCloud Management Console. Can also set access controls on users and groups of users. https://wiki.duraspace.org/display/DURACLOUDDOC/DuraCloud+Security,User roles and access rights can be managed at collection and individual file level,,Each member has their own credentials.,Yes,Yes,Yes. ,"Yes, either via internal authentication or integrated with external authentication (e.g. SAMl, LDAP, others)"
Member / customer initiated content restoration ,Members submit request to MetaArchive central staff who establish a secure connection to a member's LOCKSS server that has copies of the requested content. An uncompressed ZIP package is constructed and made available for the member to download,Users utilized Retrieval tool to download content to local environment https://wiki.duraspace.org/display/DURACLOUDDOC/DuraCloud+Retrieval+Tool. For DuraCloud Vault there is a user initiated restore from either DPN or Chronopolis. ,"Universal Access module allows users to search, find, and download publicly available content. Bulk download / restoration details not available. Many different levels of accessibility that can be implemented including user defined restrictions.",,Through the APTrust API.,Yes,they just copy files back out through fileshare. If they are using Archivematica they can get IPs out from there.,it can be. through duracloud interface. bu through TDL for S3 and Glacier,
,,,,,,,,,
Integration with other systems ,,"Can be integrated as backup for DSpace, DSpace Direct, and Archive-It. Archivematica. DPN, Chronopolis. REST API that allows anyone to write an integration. Client tooling for the REST API in Ruby and Java. ","Supports workflows to automate the bulk ingest of exported DSpace, PastPerfect, CONTENTdm, SharePoint, Outlook and Gmail package and advanced Website harvesting. Also supports integration with catalog systems including ArchivesSpace, Axiell CALM and AdLib for data management. ",,Any system since the main point of contact is the API for bags. We also create PREMIS logs that can be imported into any system.,"Archivematica (through DuraCloud), DPN","CRIS and publishing services such as EPrints, DSpace, Pure, Archivematica, and Figshare http://arkivum.com/integrations/ ",DuraCloud is stand-alone. goals for 2017 is to integrate more. Just hired services manager who will start working on strategy for integration,"Yes, OAI-PMH - publishing and harvesting, APIs, SRU and more. Integrated with LMSs such as Alma,Aleph, Voyager, Discovery systems such as Primo, Summon. Archival systems such as Ad Lib. Collection management systems. Many others.
Information on APIs and integration points is available in the Developer Network - developers.exlibrisgroup.com/rosetta.
"