This article lists the steps to migrate existing documents and their metadata into a single Meridian Enterprise vault using the Database Import Wizard in the Meridian Enterprise Configurator.
This task includes these phases that should be performed in the order listed:
- Plan the migration
- Prepare the metadata
- Run the Database Import Wizard in the test and production environments
Note This article describes the general steps required to perform a migration. The specifics of some steps will depend on the data itself, the configuration of the destination vault, and the system requirements. The specifics are beyond the scope of this article and may require additional tools, steps, and other resources to complete the migration.
Each phase of this task is described below.
Plan the migration
Although no migration plan is completely accurate, planning the document migration attempts to foresee the major migration problems so that solutions can be devised in advance of the actual migration. A plan also makes a migration repeatable and serves as a record of the activities performed to accomplish the migration.
You should write a data migration plan in advance of the migration and update it whenever new or better information is discovered.
A complete data migration plan should include (without limitation) full details of the following:
- The scope of the migration (sources and quantity of documents and metadata)
- Metadata property mappings (source property name to destination property name and data type)
- Document naming conventions and folder structures (use existing or convert to a new convention)
- Revisions and their identification
- References (how they will be identified and created)
- Document type mapping
- Workflow status mapping
- Rendition requirements and creation
When a detailed plan is complete, the source data can be prepared as described below. The preparation itself might necessitate revisions to the migration plan.
Prepare the metadata
Seldom is metadata of sufficient quality to ensure a smooth and accurate migration. There seems to always be some data scrubbing and conversion necessary before the metadata can be imported into Meridian Enterprise. Now is the time to perform all quality and consistency checks and corrections.
The metadata can originate from virtually anywhere and may be in almost any format ranging from a simple file system directory in a plain text file to tables in a database. Regardless of the original format, the first step is to convert the metadata (if necessary) into a format that can be imported by the Database Import Wizard.
The Database Import Wizard accepts many popular data formats. The most flexible of these formats is Microsoft Access. Queries can be created to perform almost any manipulations of the data that are necessary. The queries can be grouped into a macro that can automate and repeat the cleanup process through multiple test cycles. And the Access format can manage unlimited numbers of records with high performance, unlike the other available formats.
With the source metadata open in your application of choice, find and resolve any of the follow discrepancies that might exist:
- Incorrect filenames.
- Files residing in incorrect folders
- Duplicate files.
- Invalid dates.
- Records with empty fields that will be used in any part of the Meridian Enterprise folder structure, filenames, workflow status, revision numbers, or navigation views.
- Records with duplicate document numbers and duplicate revision numbers.
- Missing or unsupported versions of external reference files.
- SQL reserved characters in data dictionaries.
- Validate existing data against data dictionaries to ensure they are identical.
- Ensure that all fields required to build the vault folder structure are populated.
- Eliminate any Windows illegal filename characters (# ; % , & + \ / : * ? " < > |)
This list is by no means all-inclusive. Test the metadata migration and add any more consistency checks that are necessary.
You may also need to build (convert or concatenate) additional metadata from the original metadata in order to meet the new system requirements. Now is the time to do this.
Typical new properties that need to be built include, but are not limited to:
- Destination paths
- Destination document type names
- Destination workflow status names
- Rendition properties
- Original (legacy) document IDs and other metadata
For more information on preparing metadata for import into Meridian Enterprise, see "Creating an import source database" in the Accruent Meridian Enterprise Configuration Guide.
Run the Database Import Wizard
The hard work of migrating documents in a vault is the planning and preparation that are described above. If that work was done thoroughly and accurately, this step is comparatively easy. The Database Import Wizard does most of the work.
Before you begin
Before you run the Database Import Wizard, the server and vault must be prepared to receive the incoming documents efficiently:
- Ensure an accurate backup of the vault exists.
- Review the operating system configuration for optimal performance as described in "Optimizing the server operating system" in the Accruent Meridian Enterprise Administrator's Guide.
- Review the Meridian software and vault configurations for optimal performance as described in "Optimizing the Meridian server software", "Optimizing vault performance" and "Optimizing batch operations" in the Accruent Meridian Enterprise Administrator's Guide.
- Configure the vault for maximum batch performance as described in "Optimizing batch operations" in the Accruent Meridian Enterprise Administrator's Guide.
- Reboot the server.
- Disable all scheduled tasks.
- Review the vault status for inconsistencies as described in "Reviewing the Application event log" in the Accruent Meridian Enterprise Administrator's Guide. If inconsistencies exist, run the Vault Consistency Wizard as described in "About the Vault Consistency Toolkit" in the Accruent Meridian Enterprise Administrator's Guide.
-
Configure the vault with any new objects that are required by the migrated documents, for example:
- Document types and folder types
- Document and folder property sets and property definitions
- Reference types
- Application link configurations
- Project definitions (if required)
- Field-Path definition
- Ensure that there is sufficient free disk space on all affected volumes for the new metadata, stream contents files, and backups.
To run the Database Import Wizard:
- Perform the tasks described in "Importing documents from a database" in the Accruent Meridian Enterprise Configuration Guide.
- When you have performed a successful production migration:
- Review the vault status for inconsistencies as described in "Reviewing the Application event log" in the Accruent Meridian Enterprise Administrator's Guide. If inconsistencies exist, run the Vault Consistency Wizard as described in "About the Vault Consistency Toolkit" in the Accruent Meridian Enterprise Administrator's Guide.
- Import data dictionaries into lookup lists or tables, and create queries to external data dictionary sources.
- Initialize sequence numbers, if required.
- Perform a verified backup of the vault and the Database Import Wizard configuration file to establish a known good snapshot before the vault is put into production use.
- Thoroughly review the migrated data for completeness and accuracy. Re-import if necessary.
- Reconfigure the server and vault for multi-user use and reverse the configuration changes performed at the beginning of this task.
