Go to Top

Why should I monitor my exchange mailbox quotas to prevent data loss?

Like many of you, I have a number of routine checks that I run on the Exchange servers to keep them in good health. One of those areas is managing user mailbox quotas. I’ll often spend a couple hours a week with users to help them implement a mailbox storage diet and explain the importance of keeping their emails managed properly. However more interestingly, I have been asked numerous times how this can relate to data loss and what actions can be taken to prevent this from occurring. In this post, the second of our Exchange series, we’ll be taking a look at this in more detail.

Why should I monitor my mailbox quotas to prevent data loss?

A common error message that can appear if mailbox quotas are not managed correctly is “Email Sent Message Error: 550 Mailbox Quota Exceeded”. This can mean that either:

  1. The quota for the email account is exceeded.
  2. The quota for the domain is exceeded.

Although this error message can be remedied by the user who receives it by simply deleting items from their mailbox, this is not an ideal scenario from a business perspective.


Fig. 1. Mailbox Storage Quotas – Per Mailbox Database via Exchange Management Console [Source:  msexchange.org]

The main issue that can arise in terms of data loss is the uncontrolled growth of mailboxes to the point where it can cause an outage. Mailbox quotas allow a user and/or business to control the growth of their stores. Without quotas, there is a greater likelihood of large messages being delivered, which in turn lead to increased sizes of stores and transaction logs. If these volumes are filled to capacity, this can result in an outage and a subsequent service denial.

Other issues that can result from this are slower retrieval times when an Outlook user is searching through their mailbox for items as it must perform several operations against the Exchange server to retrieve the contents of a folder. Large message counts could also lead to more I/O on the disk subsystem and possibly cause greater failure rates on the disks or possible database corruption.

What does Microsoft have to say?

Microsoft offers advice on their support site to reduce the risk of this issue occurring, also recommending that a user maintains a range of 2,500 to 5,000 items in a folder (Exchange 2000 and 2003) for optimal performance and reduced risk of errors and outages. In Exchange 2007, improvements in I/O, larger page size, and increased cache can help enable an increase in the recommended maximum item count. With properly architected hardware, an acceptable user experience can still be maintained with item counts as high as 20,000 items. For Exchange 2010, this recommendation changes slightly, where it is recommended that you store no more than 100,000 items in core folders, such as the Inbox and Sent Items folders, and no more than 10,000 items in the Calendar and Contacts folders. And Exchange 2013 has a recommended use of 100 GB mailboxes with 1 million items. It has been noted that the Extensible Storage Engine (ESE) of Exchange 2013 has improved database performance so much that it uses 93% less I/O than Exchange 2003, a dramatic difference as can be seen in the graph below.


Fig. 2. Exchange IOPS Trend from Exchange 2003 to Exchange 2013 [Source:  Microsoft Exchange Conference 2014, Exchange Server 2013 Architecture]

The importance of mailbox quotas should not be underestimated and, as such, has been addressed quite well on the TechNet Forum for each of the different Exchange versions. Overall, their general advice for Exchange would be:

“You can help avoid poor performance in Outlook by carefully managing the number of items in folders, especially the Outlook folders that are heavily used. These folders include Inbox, Calendar, Tasks, Sent Items, and Deleted Items, with special emphasis on Inbox, Sent Items, and Deleted Items, because these are the first to be downloaded. This will also pertain to any other folders that are heavily used.”

Monitoring mailbox quotas is just one way of preventing data loss; in my next post we’ll be checking out what geo-clusters are and how they can help to avoid data loss too. In the meantime, if you’ve got any comments or feedback about this post, why not let me know in the comment box below.

See you soon!

, , , , , , , , , , , ,