NDSA:Cloud Presentations

In each case we would want to identify who would present, who will contact them. Then when they will present.

From there we can include specific questions we would like them to respond to.

Presentation Schedule and Slides

1. Feb 1, Tues, 1:00 EST call with iRods Reagan Moore (presentation)
2. Feb 14, Monday, 11:00 EST call with Duracloud (presentation)
3. Feb 17, Thurs, 11:00 EST call with MetaArchive/GDDP Katherine Skinner, Matt Schultz and Martin Halbert MetaArchive NDSA (presentation)

People/Projects to Contact

• Chronopolis (Mike Smorul will contact)
• Open questions from the Educopia Guide to Distributed Digital Preservation
• Commercial providers? (Who specifically would we want here? Please add them.)
  • Azure (Leslie to contact)
  • Amazon (Who will contact?)

General Questions for Cloud Service Presenters

Here we are working on a set of general questions for presenters to develop talks around.

1. What sort of use cases is your system designed to support? What doesn't this support?
2. What preservation standards would your system support?
3. What resources are required to support a solution implemented in your environment?
4. What infrastructure do you rely on?
5. How can your system impact digital preservation activities?
6. If we put data in your system today what systems and processes are in place so that we can get it back 10 years from now? (Take for granted a sophisticated audience that knows about multiple copies etc.)
7. What types of materials does your system handle? (documents, audio files, video file, stills, data sets, etc) And give examples of those types in practice

Responses to questions

NDSA:iRODS direct responses

Other general notes:

• [Snavely] The need for each storage target to support a specific set of operations, and consistently with other storage targets, seems like a risk that comes along with the elegant abstraction that iRODS provides. Clear specifications help mitigate this risk.

NDSA:DuraCloud direct responses

Other general notes:

• [Snavely] Treatment of cloud provider is generally as a black box, without a strong sense of actual reliability of underlying storage systems. Cloud providers tend to promise checksum validation of contents, but recourse if validation fails was unknown (right?). Additional checksum validation has been augmented on top of cloud storage service by Duracloud.

NDSA:MetaArchive/GDDP direct responses

Other general notes:

• [Snavely] Built on LOCKSS, so data integrity assurances are provided by robust networked software model augmented to commodity hardware and storage. Federated nature provides integrity assurance but also a lack of central control in that the accidental loss of multiple caches is unlikely but e.g. scheduled maintenance or upgrades could coincidentally collide.

Chronopolis

1. ...

MicroSoft Azure

1. ...

Amazon S3/EC2

1. ...

Questions for Member Institution Implementations of Large Scale Storage Architectures

1. What is the particular preservation goal or challenge you need to accomplish? (for example, re-use, public access, internal access, legal mandate, etc.)
2. What large scale storage or cloud technologies are you using to meet that challenge? Further, why did you choose these particular technologies?
3. Specifically, what kind of materials are you preserving (text, data sets, images, moving images, web pages, etc.)
4. How big is your collection? (In terms of number of objects and storage space required)
5. What are your performance requirements? Further, why are these your particular requirements?
6. What storage media have you elected to use? (Disk, Tape, etc) Further, why did you choose these particular media?
7. What do you think the key advantages of the system you use?
8. What do you think are the key problems or disadvantages your system present?
9. What important principles informed your decision about the particular tool or service you chose to use?
10. How frequently do you migrate from one system to another? Further, what is it that prompts you to make these migrations?
11. What characteristics of the storage system(s) you use do you feel are particularly well-suited to long-term digital preservation? (High levels of redundancy/resiliency, internal checksumming capabilities, automated tape refresh, etc)
12. What functionality or processes have you developed to augment your storage systems in order to meet preservation goals? (Periodic checksum validation, limited human access or novel use of permissions schemes)
13. Are there tough requirements for digital preservation, e.g. TRAC certification, that you wish were more readily handled by your storage system?

Responses to questions

NDSA:WGBH Responses

1. What is the particular preservation goal or challenge you need to accomplish? (for example, re-use, public access, internal access, legal mandate, etc.)

NYU Libraries processes, enables access to, and preserves digital materials that come from both the NYU community and from collaborating partner organizations.

2. What large scale storage or cloud technologies are you using to meet that challenge? Further, why did you choose these particular technologies?

Our current repository asset store runs on SunFire X4500 and X4540 storage servers. The data servers are mirrored and backed up to tape. We are building a new repository system using Isilon storage arrays. The Isilon arrays are mirrored, geographically distributed, and backed up to tape.

We are not pursuing cloud storage at this time.

3. Specifically, what kind of materials are you preserving (text, data sets, images, moving images, web pages, etc.)

Our preservation repository contains: - texts - images - video - audio

4. How big is your collection? (In terms of number of objects and storage space required)

Combined existing and new repository systems: 22,594 objects 81 TB (63 TB of video)

5. What are your performance requirements? Further, why are these your particular requirements?

The storage solution must be fast enough to support ongoing fixity, ingest, and access operations.

6. What storage media have you elected to use? (Disk, Tape, etc) Further, why did you choose these particular media?

We use both disk and tape (for backup). The first and second copies are stored on disk. The third copy is stored on tape.

We need content on disk because we serve some content directly from repository storage. We also transcode to create access copies served through streaming media servers.

7. What do you think the key advantages of the system you use?

The new system is under construction, but will be able to support various curation, publication, and preservation workflows. The underlying storage solution will allow us to easily add capacity to the system as needed.

8. What do you think are the key problems or disadvantages your system present?

Ingest in our current system can be rather slow due to the ingest mechanisms in our application.

9. What important principles informed your decision about the particular tool or service you chose to use?

We requested that the storage system be scalable, and ideally present a single filesystem to the applications using the storage. Our systems group then researched multiple storage solutions.

10. How frequently do you migrate from one system to another? Further, what is it that prompts you to make these migrations?

We are coming up on our first major migration in approximately four years. In addition to content in our preservation repository, we have a legacy content that is stored across multiple systems. The new repository should allow us to aggregate and manage all of our content in a single system.

11. What characteristics of the storage system(s) you use do you feel are particularly well-suited to long-term digital preservation? (High levels of redundancy/resiliency, internal checksumming capabilities, automated tape refresh, etc)

The Isilon storage system is designed to scale and includes configurable data integrity and data recovery features.

12. What functionality or processes have you developed to augment your storage systems in order to meet preservation goals? (Periodic checksum validation, limited human access or novel use of permissions schemes)

Ongoing fixity checks and "completeness" checks.

13. Are there tough requirements for digital preservation, e.g. TRAC certification, that you wish were more readily handled by your storage system?

Not at this time.

General Concerns

1. confidential data
2. encrypted data
3. auditing
4. preservation risks
5. legal compliance
6. ...

Solution Models and Environments

Name	Offered as Service	Deployed Locally	Opensource	Authentication Scheme	Ingest Mechanism	Export Mechanism	Integrity/Validation Mechanism	Replication Mechanism	Administration Model (Federated, etc.)	Tiering Support
iRODS
DuraCloud	yes	yes	yes (Apache2)	Basic Auth	1:web-ui, 2:client-side utility, 3:REST-API	1:web-ui, 2:client-side utility, 3:REST-API	Checksum verified on ingest. On-demand checksum verification service.	Built-in support for cross-cloud replication.
MetaArchive/GDDP
Chronopolis
Microsoft Azure
Amazon S3/EC2