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 although the contents are most likely out of date.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
> >
>