AnswerArchiver uses the following methods to keep the Archiver archive store as small as possible:
Single Instance Storage
If a single email is received by multiple local recipients, only one copy of the message is stored in the Archiver archive store. The archive will contain a reference to this email for each of the recipients.
MailArchiver 2012 improved this feature and follows the flow outlined below:
- An email flows into Archiver (e.g. via the journal mailbox, the import service or the Archive Assistant)
- GFI Archiver determines the email's date and finds the corresponding Archive Store for it to save it to (if none exists, is read only or flagged to not store further emails a new Archive Store is automatically created via the Archive Store Schedule)
- Before saving the email into the Archive Store GFI Archiver checks (via a hash stored for each email) if the email is already stored in the Archive Store determined in step 2
- If it does not exist, the email will be archived
- If it does exist, the email will not be archived a second time (but new owners are added)
GFI Archiver will attempt to compress a message as much as possible, before archiving it. An exact compression ratio is not available since this will depend on the contents of the message and attachments which are included in the email. For example, a message containing text will be highly compressed, while compression for a .zip file will be minimal. Attachments which are already compressed such as .jpg, .avi or Microsoft Office 2007 documents will have a low compression ratio.
GFI Archiver will only store the compressed message if the size of this compressed message is smaller than the size of the actual email. To achieve maximum efficiency when reading the emails, if there is no size difference, then the actual message will be stored to the GFI Archiver archive store rather then the compressed message.