Slide notes for presentation to Toronto Area Archivists' Group, Sept. 11th 2015. See also slides: http://www.slideshare.net/Archivematica/getting-started-with-atom-and-archivematica-for-digital-preservation-and-access-52764372
Getting Started with AtoM and Archivematica for Digital Preservation and Access- notes
● We provide installation instructions in our documentation. You will need to have a web
server, preferably running Linux although installations have been done in other
environments too. There are a number of software dependencies that you’ll need to
have installed all are listed in our docs. This installation is advanced you are likely an IT
person if you are attempting it.
● If you want to have your own AtoM installation just to test with, we also have instructions
to download and install a Virtual Box version. This installation is possible for the
relatively techsavvy. You can also use the publicly available demo site, but data is
● If you do not have IT support and you cannot install the software yourself, you might
consider a hosted version. Artefactual Systems offers a variety of hosting plans on an
annual basis. The prices are nonprofit friendly.
● Many Canadian institutions are using AtoM locally as well as contributing to their
provincial portal. Their reasons could be:
○ Greater control over the look and feel of the site, menu labels, and other
customizable aspects of the interface
○ Set your own policies for digital objects, in terms of the size you upload and how
○ Use the internal aspects accession records, donor records, make your own
static pages, etc.
● Some institutions that have their own AtoM installation send their records to Amanda to
upload to Archeion via an EAD export, which you can do out of AtoM with a click and file
● Possibly. How easy that is to do will depend on the format it comes in. In general, AtoM
can import two different kinds of data files EAD or CSV. CSV needs to be formatted in a
specific way for AtoM to be able to understand it we provide instruction in our
● If your data is in some other machinereadable format, like XML which is not EAD
formatted, then it may be possible to write a script to transform it into something that
AtoM can interpret.
● Data migration is something that you can contract us to do, or you can have
inhouse/other developers look at your data.
● We’ve built AtoM purposefully so that it can have a different theme as a plugin. A
developer would need to write the code. AtoM versions 2 and above come with two
themes Dominion and ArchivesCanada. The ArchivesCanada theme is built off of the
Dominion theme so developers can look at it and use it as a model for how to write their
own theme. Here are some examples of nice looking AtoM sites:
● The great part about being part of an open source community is you can feel a sense of
ownership over the direction of the project. We encourage participation from any AtoM
user to contribute ideas of how the software could be improved. Artefactual Systems
works on what’s called the “bounty model” of software development new features are
sponsored by a user or institution, and when the next public release of the software is
made, we incorporate those sponsored features.
● If you don’t have sponsorship dollars, you can still contribute ideas for improvements in
our user forum. We maintain a wishlist of things that the community has suggested and
when there is a good fit with a sponsoring institution, we try to “shop” it to them.
● Developments proposed by institutions are not automatically incorporated into the public
project. We archivists help steer the project to make sure that new features are generally
usable by the community at large.
● The other way to contribute is through code. If you or someone at your institution has
written custom code for an AtoM feature, you can submit a pull request. We do our best
to include these in the next public release of the software sometimes it takes a while, or
takes additional sponsors, to help clean the code up and make it appropriate for general
Notable additions to the latest release of AtoM (version 2.2):
● Generate finding aids you can create RAD compliant finding aids in RTF or PDF format,
and control whether they contain all the available descriptive fields or are a “inventory
● Create a PREMIS rights statement that controls whether or not the public can see a
digital object (or whether they can see the full size or only the thumbnail, etc).
● Dates of creation (of the records) added to the accessions template
● Some general usability enhancements better Settings page, “Part of” information
included when browsing descriptions, new search icon and revised browse menu.
AtoM version 2.3 will likely be released late this calendar year or maybe early next calendar
year. Software releases are part of our community work for AtoM they’re not sponsored by any
institution or foundation so we balance our workload to get releases out the door a couple of
times a year. Version 2.3 will have:
ability to edit title, slug and filename of digital objects through the user interface
ability to strip filenames that come from Archivematica
accessibility enhancements including better keyboard navigability
table view of institutional browse
a “clipboard” for researchers to save groups of records
an optional new treeview which takes up the same width as the description column
● Archivematica must be installed on an ubuntu web server. To install Archivematica, you
should be comfortable logging into a server, downloading packages and running
● Archivematica in a sense comes with three parts to install the dashboard, which we just
looked at, a separate application called the Storage Service, and a package of
dependencies so that Archivematica can run them as microservices.
● Folks often ask, how big of a server do I need? How much storage do I need? We have
recommendations in our documentation, but the short answer is larger collections will
require more processing space and obviously more storage space. In addition to the
space you need to store the collection, you should have at least 20 additional GBs.
● If you’re just trying it out, we have an online demo you can use; if you want to run tests
with your own material, you might consider downloading to a desktop running Ubuntu.
● We’re offering a hosted service with DuraCloud now called ArchivesDirect. DuraCloud is
part of Duraspace, an American nonprofit which is dedicated to helping libraries and
archives preserve and make available digital assets. Because DuraCloud is American,
the servers are in the States which causes some problems for many Canadian
institutions who want to use the service.
● We’re also involved in a consortial hosting systems through the Council of Prairie and
Pacific University Libraries. The hosting is done at UBC, but is only available to COPPUL
● Plans are in the works for a Canadianbased hosting solution for Archivematica. You can
expect more news on this by early next year. We want Canadian institutions to be able to
use Archivematica more easily, and by offering the system in a hosted environment, you
can worry less about the tech and more about the actual archivist work needed to
preserve digital material.
● There are different “flavours” or “levels” of digital preservation and I’m not here to tell you
that it’s our way or the highway. But the advantage of using a system like Archivematica
is that it bundles together many discrete digital preservation tasks and runs them all on
your content tasks like fixity checking, virus scanning, file format identification and
characterization, and normalization into formats that are considered friendly for longterm
preservation and access. A repository system like Fedora or Islandora can be
complemented nicely by Archivematica, allowing you to sleep a little better at night
knowing that you’ve taken additional steps towards longterm digital preservation.
● Archivematica is purposefully storage agnostic, which allows it to connect to a number of
different storage options:
○ You can store your preserved material on networked servers
○ You can use cloudbased storage like DuraCloud or Arkivum
○ You can use an object store like Swift
○ You can store packages in LOCKSS
○ Fedora storage could be experimented with at this stage
○ In the future you’ll be able to store preservation packages in DSpace.
● Archivematica comes with a Storage Service to help you manage the packages you’ve
put in storage, but deciding on which type of storage to use is outside the purview of
Archivematica’s activities. It’s most likely to depend on your institutional environment.
● We consider AtoM to be the “default” access system for use with Archivematica, but
other systems that have some manner of integration include CONTENTdm, Islandora,
DSpace, Archivists’ Toolkit, and soon, ArchivesSpace.
● You can also make packages of material for access and store them/download them and
manually upload them to other systems if you desire.
● Our philosophy is, your preservation system shouldn’t dictate how you provide access to
your digital material. We believe in creating many possible connections between the
software in your arsenal we like to think of it as “handshaking”.
● This process can also work the other way around you may have an institutional
repository such as DSpace into which students and faculty deposit material directly, and
then you can use Archivematica to ingest that material as a preservation backend.
● The degree of automation is largely up to you. Most of Archivematica’s processes can be
automated by changing the administrative settings in the web interface.
● Further automation can be achieved through special scripts that we call “automation
tools”. We have users who have Archivematica running in the background on materials
with literally no human intervention unless there is an error to troubleshoot. To use these
tools, you would need to alter the scripts for your particular situation the scripts can do
things like look for particularly formated filenames, use metadata csv files to make
matches between metadata and digital objects, etc.
● I think of automation like this: the more uniform your materials are, the more likely you’ll
be able to take advantage of automation techniques to preserve them. For example, if
you’re digitizing photographs and they’re all the same format, and you have control over
how they’re created, that’s a fairly easy and automatable preservation use case.
Borndigital material, which you have little to no control over, is much harder to automate
workflows for. Many institutions have a mix of both of these types of use cases and
many things in between.
● The best thing you can do as a start is analyze the content you need to preserve. Find
out everything you can about it where is it stored, what are the file formats, what kind of
media is it stored on, what contextual information does it have accompanying it, how
large in terms of number of files and storage space if possible.
● Start with some of your easier use cases. Digitization projects are great examples
because they’re relatively low risk you created the files, you can create multiple copies
so that you can experiment with preservation workflows without major consequences if
you lose material accidentally.
● If you think Archivematica is a useful tool for you, watch some of our webinars on
Youtube and read up in the documentation. Once you have installation figured out, you
can start running test packages through the system. Again, start with something easy,
understandable and low risk, and work your way up to more complex workflows.