How to Create an Archive Database in Exchange Server 2010

Exchange Server 2010 has built-in archiving features to help manage user mailbox sizes.  The Personal Archive is essentially a second mailbox assigned to the user.  Mailbox content is then archived manually, via rules, or using Retention Policies.

The Exchange 2010 archive mailbox can be located in the same mailbox database as the user’s primary mailbox, however Exchange Server 2010 SP1 added the option to store the archive mailbox in a different mailbox database.

There are several benefits to locating personal archive mailboxes in a dedicated archive mailbox database.

  • Maintain smaller mailbox databases for primary user mailboxes, allowing for faster backup and recovery of the more important mailbox data
  • Exclude archive mailboxes from Database Availability Group replication
  • Use a different backup schedule for archive mailbox data
  • Consolidate archive mailbox storage to a single database or single Mailbox server
  • Store archive mailbox databases on separate storage

Although some people might think that archive mailboxes can be stored on cheaper, slower storage, remember that Outlook clients do not cache the archive mailbox and so the I/O load on the archive database could potentially be very high if it is being accessed by a lot of users simultaneously.

Creating an Exchange Server 2010 Archive Database

To create an archive mailbox database in Exchange Server 2010 start by simply creating a new mailbox database.

Create an Exchange Server 2010 Mailbox Database

Create an Exchange Server 2010 Mailbox Database

Next, use the Exchange Management Shell to disable automatic mailbox provisioning on the new database.

[PS] C:\>Set-MailboxDatabase "Archive Mailboxes" -IsExcludedFromProvisioning $true

This will prevent the mailbox provisioning load balancer from choosing the archive database when creating new user mailboxes.

About Paul Cunningham

Paul is a Microsoft Exchange Server MVP and publisher of Exchange Server Pro. He also holds several Microsoft certifications including for Exchange Server 2007, 2010 and 2013. Connect with Paul on Twitter and Google+.

Comments

  1. Ray Bohanna says:

    Thank you this was very helpful.

  2. Chris Wightman says:

    This worked perfectly, also thanks for the “disable automatic mailbox provisioning on the new database” reminder.

  3. Why would you not want an archive mailbox database to be part of a DAG? Are there operational problems with doing that? I would have thought a DAG would offer resiliency?

    • Thats a business/technical decision for you. Some customers may not require HA for archived email.

      Others may choose to include it in the DAG but have fewer copies, eg 4 copies for primary mailbox databases, but only 2 copies for archive mailbox databases.

      The point being, having a separate database dedicated for archive mailboxes gives you that flexibility if required.

      • Thanks Paul. I was worried that you couldn’t have them as part of a DAG! It’s nice having them in a seperate database for the reasons you mention. Nothing worse than working late trying to move massive mailboxes!!

  4. Hi Paul,
    I’m wondering if it is possible to add another exchange server with only the mailbox role just for archive databases without having to create a DAG? Our current environment is one exchange server (SP3) with all roles. I can’t find any info about adding another server and not having it be part of a DAG.

    • Yes, you can have as many standalone mailbox servers as you like. Just install the mailbox role on its own. There’s no requirement to create a DAG to have multiple mailbox servers.

      The only requirement is that it must be in an AD Site where at least one CAS and HT server also exist.

  5. Darrell Q says:

    Hello Paul,
    I’m looking into this for our small organization right now, to help with our space management issues.
    We have a 2 member dag, with 5 250GB databases replicated between the 2. Approximately 600 mailboxes between the works.
    Could we simply create 1 big archive DB, say 3 TB, and replicated between the 2 DAG members?
    Basically, we have access to a bunch of disk, and want to get our mailboxes back to a more uniform size of 1.5 GB or so. With the excess being auto archived to the archive DB.

Leave a Comment

*

We are an Authorized DigiCert™ SSL Partner.