What is Exchange Online Archiving?
I’ve been meaning to write this article for quite a while now, so I’m glad it’s finally “ready”. First, let me start by introducing what Exchange Online Archiving (EOA in short) actually is.
This feature, first available since Exchange Hybrid, allows you to provision an cloud-based archive for an on-premises mailbox. While having an Exchange archive isn’t something new, at least not since Exchange 2010, the fact that the archive doesn’t have to be hosted within your own organization is pretty interesting.
Archives can be useful in many ways. One of the primary reasons why archives are used is to keep historical data for a longer period of time without cluttering a user’s primary mailbox. This could, for instance, be the case when you have to meet some compliance requirements which e.g. state that corporate data should be kept for 5 years. Although Exchange doesn't have a problem with handling very large mailboxes including a high item count per folder, it’s usually the human component that cannot handle the overload of information that comes with having large amounts of data – at least that’s my experience. Keeping email inherently means that you’ll have to increase disk space to support the sometimes huge amounts of data that is involved. Although disk space has become quite cheap and Exchange 2013 is a great candidate to be used in combination with those cheap disks, there’s still a significant overhead involved in keeping that additional piece of infrastructure up and running.
This is where Exchange Online Archives could come in handy. First of all, there is no feature difference between an on-premises archive or a cloud-based (Office 365) archive. From a user’s point-of-view they both act and look the same. In fact, you are only offloading the task of storing archives to Office 365. The Exchange Online Plan 2 subscription automatically includes the right to provision unlimited-sized archives for your users. Although I don’t expect many people to run into the issue of filling up the initial 100GB, which you get provisioned to start with, any time soon, it’s very hard to match that offer for only 8$ per user per month… If you are only interested in EOA, there are specific EOA licenses as well which cost only a fraction of the full Exchange online license. Of course, this license will only allow you to use EOA and nothing more.
How does it work?
As briefly touched upon earlier, being able to use Exchange Online Archives is a by-product from having a hybrid Exchange deployment. A hybrid deployment, as the name stipulates, is the process of ‘pairing’ your On-Premises Exchange organization to Office 365; essentially creating one large “virtual Exchange organization”. As a result, having a (fully functional) Hybrid Deployment is the first requirement to abide to… Technically speaking it would be possible to setup a sort of minimalistic Hybrid deployment in which you leave out functionalities that you do not necessarily need to make Online Archives work (like e.g. cross-premises mail flow). Nonetheless I strongly encourage to still setup the full monty. It might save you some time afterwards if you decide to deploy cloud-based mailboxes anyway.
A very import part of the setup is set aside for DirSync. As you might remember, if you tick the “Hybrid Deployment” checkbox during DirSync setup, you allow it to write back some attributes into your on-premises organization. One of these attributes is the msExchArchiveStatus attribute. This attribute is a flag telling the on-premises organization whether an online archive has been provisioned or not. As we will see later in this section, this attribute is particularly important during the creation of an archive.
One of the questions I get asked regularly is whether you are required to deploy ADFS when setting up a hybrid deployment. The short answer is no. On the other hand, there are many good reasons why you would want to deploy ADFS, or rather: there are many good reasons why you would want to have some sort of single/same sign on. One reason I can think of it to simplify using online archives from an end user’s perspective. That way they won’t need to manage another set of credentials. Of course this isn’t only valid for online archives, it’s the same for each cloud-based workload in Office 365. ADFS can be one way of providing SSO, Password Sync is another. Both are valid options, neither are required and won’t be discussed here.
From a functional point-of-view, Online Archives have the exact same requirements as on-premises archives. You at least need Office 2007 SP3 Professional edition or later. Since we are running archives from Office 365, you also need to make sure to be up to speed with the latest required updates. For more information on what updates are needed, have a look at the following web page: http://office.microsoft.com/en-us/office365-suite-help/software-requirements-for-office-365-for-business-HA102817357.aspx
Now that we got the prerequisites covered, let’s have a look at how the provisioning process works from a high-level perspective:
As you can derive from the image above, there are two DirSync operations needed. The first one is used to “tell” Office 365 to create an archive for user “X”. The second DirSync operation is used to sync back the msExchArchiveStatus attribute which will now have a value of 1 instead of 0. This is to tell the on-premises organization the archive has been created. A good way to verify whether this process has completed is to run the Get-Mailbox | fl *arch* command:
Here you can see that the archive was created successfully (ArchiveStatus = Active). However, we are missing a part of the information. This is because the on-premises organization cannot provide the information from Office 365 (which is essentially another Exchange organization). To fetch the missing information, you’ll have to open up a remote PowerShell session to Exchange Online and run the Get-MailUser | fl *arch* command:
This is it for part one of this article.
In the following part, I will talk about some of the gotchas, do's and don'ts. Stay tuned!
Michael Van Horenbeeck