After end to end testing is completed, it is important to evaluate and configure acceptable threshold values for extraction and ingestion failures. Doing so will prevent the review of every PST with one or two failures and permit an Operator to focus on larger areas of concern.
Once configuration and testing have been completed, the Migration Agent will need to be packaged and tested for deployment. This is frequently done on a small group of users close to the Migration Manager for PSTs team. If the deployment installation went as expected and the workstations can be seen reporting back to the Migration Manager for PSTs server, the system will need to be less aggressively configured to accommodate the added communication from the agents.
At this point you can begin the deployment of the agents to the remaining workstations in an environment. The business pilot will consist of the users who tested the agent package deployment and will provide an opportunity to begin to see bottlenecks within the environment. Tuning the system to a balance of resource consumption and performance is frequently initiated during the business pilot because it is the first time in a project that the system could be busy enough to maximize the performance of the system. Other pilots may wish to be performed depending on the results of initial pilot attempts.
After deployment of the Agent to the remaining workstations occurs, the discovery of PST files begins. It is suggested to let this process run until the number of discovered PSTs starts to level out. In smaller projects this may take three or more weeks. Larger projects may take longer.
During the discovery process and through its completion, it is suggested that Operators focus their efforts on ownership identification through the options within Manage > Owners. Accurate and complete ownership identification is one of the challenges in PST migration projects. Migration Manager for PSTs is able to provide suggestions as to ownerships immediately following the initial discovery results being returned. Aggressively managing ownership of files during the discovery phase of a project will drastically reduce the challenges and time needed to manage file ownership at later phases of the project.
The number and volume of PST files may continue to grow as your migration continues, but when discovery is nearing the end, consistently sustained growth can be seen week over week. At this point waves of users can start being prepared for migration. Typically, this initially involves ensuring communication has gone out to impacted users and enablement groups have been defined for the migration.
The Ramp-up phase implies growing the wave sizes until waves are keeping the system busy with available disk space and mostly completing them in the desired wave interval. Setting the Migration Priority for a wave of users enables them for migration. Commonly, a unique priority value would be set per migration wave. Daily monitoring is required at this point of the migration. Monitoring should consist of the environment, the active wave(s), and those users’ files until mostly completed. Remediation and reprocessing of some PST files or items may be required for the wave to fully complete. In addition, module tuning to achieve more performance within the limitations of the environment should also be performed during the initial monitoring and growth of migration waves.
You do not have to wait for a wave to be entirely completed prior to starting the next wave. The goal of an efficient migration project should be to minimize backlog but ensure enough work to have the environment persistently working. This typical bottleneck in a migration work flow is typically the extraction module but could be other areas if certain environmental factors dictate it. Keeping a system busy requires persistent uploads and consistent ingestion to ensure there is enough work to keep the system working at its fullest capacity without running out of available disk space.