Outlook Personal Folders (PST files) have been a management challenge for both small and enterprise companies for a number of years. With the industry and business culture shift toward eliminating PST files, companies are faced with the burden of moving this data in a timely, secure and cost-effective manner. Sherpa Software is a trusted advisor and solution provider that helps companies through their PST migration process. Our experience in planning, implementing and executing the migration process has proven successful in addressing challenges these companies face. This case study outlines one of our many successful PST migration projects.
A healthcare manufacturing organization contacted Sherpa requesting help in accomplishing their data migration objective: to migrate a vast amount of PST data into a centralized data store. Sherpa’s Mail Attender solution has the capability to migrate PST data to target stores that could include mailboxes, online archives, Office 365 mailboxes or Office 365 archives as well as other custom targets. This healthcare manufacturing company had 6,000 employees and 50,000 PST files with 150TB of data spread across three sites.
When the organization first contacted Sherpa, one of their main challenges was that they had no idea of the number of PST files in their environment, nor did they know where these PSTs were located. They could only tell that the PST files might reside on desktops, laptops or network shares. The company was looking for a solution that could do the following:
- Inventory their PST files, regardless of location
- Analyze how much data needed to be ingested
- Determine how long the ingestion process would take
- Automate the ingestion process
- Ensure that data was migrated into the correct user’s online archive
- Filter data to ingest only the last three years’ worth of email
- Provide detailed audit reports
Sherpa’s Mail Attender product was the perfect fit, and the organization decided to move forward with the evaluation process.
During the evaluation, the company used Mail Attender to run migration scenarios in a test environment. They used sample PST files to run simulations of their project requirements. Testing was successful, and planning was quickly underway for a full migration deployment.
Phase One: Planning and Production Testing
Once the organization verified the PST migration test data, key questions needed to be answered, including how much data would be migrated, how long the project would take and how the deployment would impact their infrastructure and end-users.
The organization began by deploying Mail Attender’s desktop agents using a login script. This locally installed software allows the centralized Mail Attender console to “talk” with each desktop and laptop in order to both identify and process the PSTs stored on end-user machines.
The company also initialized automated searches against their file shares to find any network-based PST files. Between the network shares and desktop agents, Mail Attender discovered 50,000 PST files, totaling 150 TB worth of data. This reflected an average of approximately eight PST files per user.
Parallel with the identification of the PST files, the company selected employees for production testing of the migration. The information collected from these pilot users was helpful in determining the impact of PST processing and benchmarks for expanded deployment, as well as identifying information required for audits and any changes needed for the migration process. It was confirmed that the end-users saw little to no impact during processing, and only a few tweaks were made to the ongoing process.
Thanks to the data collected during this test, it was determined that it would take less than 60 days to process all 150 TB of data. The organization was thrilled as it would result in a quick ROI for the project.
Phase Two: Migration Preparation
Once pilot testing was finalized, the company leveraged the reporting functionality of Mail Attender to ensure the information required for auditing purposes was being captured. For each PST file that was processed, the report included the following:
- Name and location of the PST file
- Name of the AD user who ‘owned’ the PST file
- Total size (volume) of the data
- Total number (count) of messages
- Total size (volume) of the messages that were less than three years old
- Total number (count) of messages that were less than three years old
This information benefited the organization in a number of ways. It provided a detailed background of the data, which allowed the company to plan for growth of the destination data stores once the ingestion was complete.
Additionally, these reports were used to help determine what data could not be processed. Any data migration will include a minority of outlying data which cannot be processed; the reasons for this may include corrupt files or an inability to access a PST file. Determining which items were unable to be processed and why is essential for both auditing and troubleshooting purposes. To further identify these items, Mail Attender was used to create post-processing reports for the destination data store, which included:
- The name of the AD user who ‘owned’ the data store
- The total size of the folder where the PST data was ingested
- The total number (count) of messages in the folder where the PST data was ingested
Through Mail Attender’s reporting details, it was discovered that a handful of users were on leave during the information collection stage; because of this, the desktop agents were never installed on their machines. Thanks to the various console views in Mail Attender, these users were easily identified, which resulted in manually installing the desktop agent on the missing machines and allowed their data to be processed with the others.
In uncovering the missing machines, the organization wanted to further leverage Sherpa’s expertise through a professional services engagement. Sherpa’s Professional Services team provided the organization with the manpower and experience to perform the migration.
The PST files that needed to be processed were ready to be ingested into the appropriate target data stores. The organization knew that manually assigning users to each PST file would be a cumbersome task to perform for 50,000 PST files; as a result, Sherpa recommended the Scan for Ownership process in Mail Attender to easily automate this requirement and save numerous hours on the project. The Scan for Ownership process allows Mail Attender to analyze the file to determine which AD user “owns” the PST.
During this process, Sherpa successfully identified a small number of PST files that did not have owners assigned because they were either former employees (therefore not in AD) or created from a service account that did not need to exist. After locating these PST files, the organization determined this data did not need to be migrated into Exchange.
Phase Three: Migration
Once all of the PST files identified for migration were ready for processing, the end-users were notified to stop using PST files and anything “new” was not going to be migrated. The report process was run one final time for up-to-date statistics, and the migration began.
To make this process more manageable, the ingestion was broken up into three groups based on the user’s name (A – J, K – R and S – Z). The first group was processed during the first week with the second and third groups following in two-week intervals. Each run took a few minutes to setup and initialize.
While the groups were processing, Sherpa’s Professional Service team regularly checked that ample space remained on hard drives for the ingestion process to successfully run. Additionally, Mail Attender was configured to notify the Sherpa technicians of any PST data that was not able to be processed.
These notifications included specific information on why PST data was not able to be processed to enable the troubleshooting process to begin immediately.
By the end of the ingestion, about 50 error notifications were received, most of which were regarding specific message corruption within a PST file. There were also a few messages that were identified by anti-virus as threats which were promptly quarantined. Additionally, several PST files were located on an encrypted hard drive and could not be processed. Each of these errors took less than 15 minutes to identify and troubleshoot.
During the subsequent weeks, reports were run against the destination store to verify that all data was properly ingested. Of the 50,000 PST files that were processed, less than 50 PST files had issues during the migration. In case further investigation would be warranted, Sherpa provided the organization with PST files and individual message failures, as well as final reports reconciling the PST files and the target data stores.
In a short period of time, the healthcare manufacturing organization was able to evaluate, test, plan, implement and complete a full PST migration project. Thanks to Mail Attender and Sherpa’s Professional Service team, this process was automated and efficiently executed. Because the organization was so impressed with Mail Attender, they decided to further utilize its functionality to apply retention policies against their mailboxes and archives.
Since this migration, Sherpa has further improved the migration process by adding new features to Mail Attender. Among other features, the ownership assignment has been improved as it can now be based on the logged-on user. This assignment increases the speed of determining the PST owner since no data analysis is necessary. Furthermore, Mail Attender can now be configured to allow the desktop agents to make the AD queries during the Scan for Ownership assignments, or pass the information along to the console machines and have them perform the AD queries. This is ideal for companies who have more restrictive environments, and do not allow AD queries to be performed on the desktops.
Finally, Mail Attender now has the ability to deduplicate the data being ingested, which can save a considerable amount of costly space. This deduplication ensures that if a message in a PST file already exists in the destination store (be it a mailbox, online archive, Office 365 mailbox or Office 365 archive), its duplicate will not be ingested. In addition, administrators have the option to control the deduplication at the folder level, or at the store level.