Gone are the days when a document migration was as simple as running a file dump and a script to load the new repository. Spending four to six weeks creating the scripts to custom load your document collection and then several trial and error attempts to move thousands, or at most a few hundred thousand, documents wasn’t trivial but was common for most migration. Today’s migrations are usually about billions of documents being stored in several different sources and sometime being moved to several destinations. Yesterday’s migration solutions don’t address today’s migration challenges.
Discovery and Mapping
Migration starts with mapping and often involves several different sources. This mapping needs to be reviewed by affected parties to ensure its accuracy. A detailed review is often performed to ensure that every document type is being addressed. Certain keywords may need to be set or calculated for those that change from source to destination platform or were not in the source platform.
With Simflofy this is easy. After pointing to a source or destination repository, the system performs a discovery. The discovery returns every document type and keyword in the system in a readable format. It also creates a table of the document types in both the source and destination repositories that were used for mapping. Mapping is done by type and involves selecting a source and destination keyword pairs. If new keyword values are needed, the system supports simple logic to generate those values. Simflofy even allows for easy integration of 3rd party tools for tasks such as categorization, classification, or analytics.
Text Scripts, Not Test Migration
Usually a test migration involves testing a subset of documents, but Simflofy treats migration like any standard quality assurance (QA) program. Once the mappings have been completed, a test is run set against the migration. Simflofy reports any errors that might exist. The test script will ensure that expected keywords will migrate appropriately.
Simflofy is both a migration and a federation platform. There’s no need to worry about migrations failing and needing to roll-back the migration.
“New” Document Migration
Simflofy migrates each document individually. One document may “fail” during migration but not the entire job. If an error does occur on a document, Simflofy can pause a migration or let the migration go on as you fix the errors. When the first migration pass is completed, Simflofy knows which documents did not migrate due to errors. This means that you can easily fix errors and run a migration refresh. And since Simflofy keeps track of changes to documents, you don’t need to disconnect users from your production system. You simply continue to run migration updates until all changes have been refreshed.
This is why the Simflofy migration approach completes migrations faster than most other tools. Traditionally most large migrations are scheduled during non-peak hours to ensure that network performance isn’t adversely affected. Simflofy can be scheduled to slow down migrations during peak hours and speed up during down time to take advantage of network bandwidth. Migrations can also be run as several projects and chained together.
The migration job can perform de-duplication and identify personally identifiable information (PII) detection. De-duplication looks at each file and its keyword to calculate the odds of being a duplicate. PII detection can indicate if a file includes sensitive information like tax ids. Third party tools can be added to the migration process. For example, tools to extract additional information from the documents or create a PDF file.
All of this makes Simflofy the fastest migration platform available today.
Simflofy can migrate billions of documents from …
Alfresco, Box, DocuWare, Dropbox, Google Drive, Hadoop, IBM Content Manager, IBM Filenet, Microsoft Sharepoint, Microsoft Sharepoint Online, OpenText Content Server, OpenText Documentum, OpenText LiveLink, Salesforce, Xerox DocuShare, and even FTP and the file system
… to any of those repositories. Or this same platform can be used to federate these repositories and allow them to work as one.