Hi Kyle,

Archivematica runs on Linux operating systems, typically configured as a
Virtual Machine on a server (but also possible to be directly installed on
a server) and users access Archivematica through a web-based dashboard.
That dashboard can be accessed by any system that can run a modern
browser. Here's
a high-level technical architecture document
<https://www.archivematica.org/en/docs/archivematica-1.9/getting-started/overview/technical/#technical-arch>
.

I second John's advice for joining the Archivematica google group!

And for John's advice on "make sure you have someone very tech-savy
involved who's comfortable with debugging." -- I think that's appropriate
advice for any organization seeking to initiate a digital preservation
project. Digital preservation is a lot of work and requires many skillsets!
No turn-key, one-click solutions for such a major operation / field of
study.

I can't say I agree with the advice for sticking to one version and not
upgrading. In my experience, organizations that don't upgrade tend to have
more problems that go unresolved and have more bugs to battle. Some people
may prefer to wait a bit after a new release in order for a patch release
<https://wiki.archivematica.org/Release_Notes> to arrive and then update.
Recent Archivematica updates have produced a much more streamlined and
faster product than previous editions, thanks to community funding.

To get to your second question, Archivematica is preservation planning
software and can store the AIPs or DIPs wherever you like. It strives to be
system-agnostic in this way -- some people store on their local servers,
some in S3, some in Fedora, et cetera. There is also mechanisms for
creating and uploading DIPs to other systems -- AtoM, ArchivesSpace,
CONTENTdm are all supported options, among others. Archivematica comes
paired with another web-based dashboard service called the Storage Service
<https://www.archivematica.org/en/docs/storage-service-0.14/> that can
manage your AIPs and DIPs -- and many users do -- but they can also be
managed by other systems. Archivematica is highly customizable and able to
fit into the unique workflows of the wide variety of institutions within
the cultural heritage sector, which is why there are many possible paths
for how it can work for you.

Ashley


On Fri, May 17, 2019 at 2:01 PM Kyle Breneman <[log in to unmask]>
wrote:

> Dear John,
>
> Thanks for your detailed email.  I'm still rather fuzzy on two points, and
> am hoping you can clarify.  First, it sounds like Archivematica is a server
> application (leaving aside the database, which obviously needs to sit on a
> server).  Its quite clear from the documentation that running Archivematica
> in a Windows environment is not supported.  I'm a little fuzzy on the
> component parts here: database on a server, server-side install of
> Archivematica, local Archivematica client running (in Linux) on users'
> PCs.  Have I captured all of the pieces correctly?  We only have Windows
> servers, so we'd need to get a Linux server up for this, right?
>
> Second, you wrote a bit about AIPs and DIPs.  Do I understand you correctly
> that both AIPs and DIPs need to be stored somewhere?  That Archivematica
> creates these packets or packages, and then passes them into some external
> system, such as Atom/ArchivesSpace/ContentDM, etc.?
>
> Regards,
> Kyle
>
> On Thu, May 16, 2019 at 2:47 PM John Pellman <[log in to unmask]>
> wrote:
>
> > 1) Archivematica does indeed use a MySQL/MariaDB database to store data
> > related to individual microservices / tasks that are performed within the
> > preservation workflow.  The MySQL database also contains the contents of
> > the format policy registry.  Specifics about what the database contains
> can
> > be found here
> > <https://wiki.archivematica.org/MCPServer#Database_Schema_Diagram>,
> > although the contents are most likely out of date.  To the best of my
> > knowledge, the MySQL database currently needs to reside on the same
> server
> > as the rest of the Archivematica installation, although I believe that
> > Artefactual (Archivematica's principal developer) had ambitions to
> decouple
> > it so that the database and Archivematica could live on separate servers.
> >
> > 2) Archivematica places an Archival Information Package (AIP) in some
> form
> > of datastore, which can be managed through the storage service console.
> > Multiple backends are supported, but my group has only ever used
> filesystem
> > storage.  Dissemination Information Packets (DIPs) are ideally
> distributed
> > via Artefactual's atom <https://www.accesstomemory.org/en/> (caveat,
> this
> > is coming from my memory to wit, which might be off; we don't use atom in
> > our library).  Archivematica/atom together are supposed to implement the
> > OAIS
> > model <https://en.wikipedia.org/wiki/Open_Archival_Information_System>.
> >
> > My advice regarding Archivematica would be to make sure you have someone
> > very tech-savy involved who's comfortable with debugging.  There's a lot
> of
> > technical debt with the project that can be challenging to deal with
> > without an IT person or a programmer (unless you hire Artefactual to set
> it
> > up for you).  I'd take a look at the user forum
> > <https://groups.google.com/forum/#!forum/archivematica> as well before
> > making a decision of whether or not to adopt it.  If you do decide to use
> > it, I would choose a version, install it, iron out all the pain points,
> and
> > then stick with that version unless there's a super compelling reason to
> > upgrade.
> >
> > On Thu, May 16, 2019 at 10:49 AM Kyle Breneman <[log in to unmask]>
> > wrote:
> >
> > > Greetings, friendly c4l crew!
> > >
> > > We're considering using Archivematica and I have several questions to
> > which
> > > I cannot find answers.  Hopefully one of you can help me out.
> > >
> > > 1) It seems like the individual Archivematica client depends upon a
> MySQL
> > > database mounted on a server.  Is that right?
> > >
> > > 2) I gather that Archivematica performs several important digital
> > > preservation tasks, but it doesn't seem that Archivematica is intended
> to
> > > be the final destination for the materials themselves.  Is the idea
> that,
> > > at the end of the workflow, what they call Dissemination Information
> > > Packets (DIPs) are routed into a separate repository, such as
> > > ContentDM/DSpace/ArchivesSpace?
> > >
> > > Regards,
> > > Kyle
> > >
> >
>