One of the tools available for account mapping in large migrations is parent and child jobs. In this technique, you run an analysis (account map generation) on a large chunk of your source data or even the full set. That map will contain all of your source and target users and groups, but it may be too large of a job to be practical for transfer purposes. You then create several child jobs that each contain a subset of the folders in the parent job. Then, rather than running separate map generations on the child jobs, you can download the generated map from the one parent job to all of the children. This can save a lot of time in running the total migration.
Parent-child jobs are also useful for when an account map is run on a full set of source accounts, but then some of those accounts are later deleted to downsize the overall account. You can run a child job on the subset of accounts that still remain, and CFP will ignore the source accounts that are no longer there, with no need to re-run and edit the original map.
The parent map can be updated if, for example, more target users are added over time, and the updated map will automatically be downloaded to all child jobs that it is linked to.
When setting up parent child jobs, do not run an account map generation on jobs that will be the child jobs. Jobs will use their own task data as the default, and will only look at uploaded parent job task data if the child job task data does not exist.
Setting Up Your Jobs
1: Set up the parent job as an account mapping job. Run an account map generation job for it. You may select multiple accounts or multiple folders in a single account.
Download the CFP generated spreadsheet and complete all of its tabs as you would like it to be implemented for all child jobs.
2: Upload the edited spreadsheet and ensure it validates. While this step is optional on the parent job, it can be helpful to eliminate errors on the child jobs later on.
3: You now need to decide how to split up your job. Go to the Analysis Page of your map run. The Users and File System tabs on the Analysis page are generally the most helpful in assessing how to split up your job. Ideally, jobs should be no more than 1 million files each, but with sufficient resources (particularly for Windows sourced jobs), CFP can generally run up to 2 million files per job. Additional factors, particularly dense folder sets, may decrease the maximum recommended size for each split of your job.
4: Once you know how you will split up your job, you are ready to set up the child jobs.
1: Create an account mapping job. Be sure to select the same source and target systems as were used for the parent job.
2: Select the desired subset of data for this particular job from the Accounts and Files page. Alternatively, if you have a large number of accounts, you can work off an exported .csv file of the source data (accounts or directories) and simply slice that .csv up into smaller pieces. Each piece can then be imported into a single child job.
3: When you get to the Settings page, select a Parent job from the Parent Job dropdown:
Only valid selections will be displayed in the dropdown.
4: Upload the edited spreadsheet. You can use the edited spreadsheet from the parent job, or you can use a smaller subset of the parent job, provided it has all necessary directives for the child job. For example, you can delete paths from the Paths tab that do not apply to the selected data set for this particular job.
5: Proceed as usual with running a simulation and transfer.