PST Migration Case Study46,000 PST files.
No manual handling.

Organisation

US Department of Education

Sector

Federal government · USA

Source

PST files on 16 file servers

Destination

Office 365

Scale

46,000+ PSTs · 6,000 users · 22 TB

Footprint

3 data centres · 10 regional locations

The situation

The department had already finished the main event: its Exchange mailboxes were in Office 365. What remained was the long tail. Over the years, staff had saved mail into PST files, and those files had spread across file servers in data centres and regional offices. There were more than 46,000 of them, holding 22 TB of email for around 6,000 users.

Moving PSTs by hand is a familiar routine: find the file, work out whose it is, rename it, copy it somewhere central, upload it, map it to a mailbox. At a few dozen files, that's tedious. At 46,000 it's impossible to do accurately, and an inaccurately assigned PST means one person's email in another person's mailbox.

How the platform was chosen

Microsoft shortlisted the platform and introduced it to the department's outsourced IT partner, who selected it. The deciding factors were a fully automated workflow from discovery to import, an architecture that could run at each regional site, granular reporting, and support for Enterprise Vault in case archives turned up alongside the PSTs.

How the project ran

1. Discover locally, not over the WAN. Agents were deployed to the regional sites. Each one found and analysed the PSTs near it, so gigabytes of files didn't have to cross wide-area links just to be inspected.
2. Assign ownership by rule. Discovery gathered ownership evidence for each file. A set of rules turned that evidence into an owner automatically, with no one opening files to guess.
3. Clean and prepare. Each PST was collected, compressed, given a unique name and had corrupt items removed, so a damaged message couldn't break an import later.
4. Feed Microsoft's import service. The department had chosen Microsoft's drive-shipping service to get 22 TB into Office 365. The platform staged the prepared files at each site for shipping and generated the import mapping file that tells Office 365 which PST belongs in which mailbox.

Results

• 46,000+ PST files discovered across 16 file servers, 3 data centres and 10 regional sites
• Owners assigned automatically by rule, for around 6,000 users
• Corrupt items removed and files uniquely named before import
• The Office 365 import mapping generated automatically
• A fully automated process from discovery to import, with no manual file handling

What this means for your project

Two things made this work. Analysis happened next to the data, at each site, so the network was never the bottleneck. And ownership was decided by evidence and rules rather than by people, which is the only way to get it right at this scale. Whether your PSTs go to Microsoft by network or by drive, the hard part is the same: knowing what exists and whose it is.

Related