A new data migration project can be a seemingly challenging process. We put our experience into the below checklist in the hopes that it will help you with your next data move. Feel free to suggest your own tips in the comments.
Data migration checklist
Step 1: Pre-migration project plan
Pre-migration impact assessment: Enthusiasm is to be admired, but data migration should be carried out in as systematic a manner as possible. Here focus on answering questions such as:
- Is this project viable?
- What is the timeframe?
- What deadlines exist?
- What resources are required?
- What software is needed?
- What hardware is needed?
- How will this disrupt the business? Will there be downtime?
- What are the risks and costs?
- Initial Next Steps: Make stakeholders aware of the pending tasks.
- Are stakeholders aware of the benefits?
- Are stakeholders aware of key dates and activities?
- Have all decision makers signed off on the project?
- Have resources been effectively allocated? Are they aware of their roles?
- Is the optimal project delivery structure defined? Does it include highly focused milestones?
- Have you created a structured workflow with tasks, dates, and deliverables?
- Are documentation and the training plan in place?
- Is the collaborative environment/platform in place?
- Have you shared the stakeholder communication plan and directory with the team?
- Have well-defined project policies been shared?
- Does the high-level project plan exist?
- Have you begun to analyze the data?
- Have timelines been refined after data analysis?
- Have standard project docs and templates been shared with the team?
- Have 3rd party agreements and requirements been formalized?
- Have future hardware and software requirements been shared with the team?
- Has the data dictionary or project wiki been created?
- Has the high-level source to target mapping been specified?
- Has the data been sufficiently pruned, and the scope defined, to result in a fast-loading database?
- Has risk discovery and analysis been revised?
- Are users aware the old system is being retired? Is a strategy in place?
- Have all modeling layers being created and revised?
- Have estimates been adjusted for new information?
- Has the detailed mapping design document been completed, reviewed, and understood?
- Have all sources and targets, down to the attribute level, been defined?
- Have interfaces for data extraction and loading been designed and prototyped?
- Have the data quality management routines been specified?
- Is the plan ready to go to the developers?
- Has hardware been specified?
- Are all service level agreements in place?
- Has the migration logic been documented?
- Has the migration been tested on a cloned staging set-up?
- Is the data validation routine or engine ready to go?
- Will data quality be monitored in an ongoing fashion? It should be!
- Is there a migration fallback system in place?
- At this point all strategies should be defined.
- Are you keeping an accurate log for compliance?
- Has the migration been independently verified?
- Have all system retirement validations been completed?
- Once all preconditions have been met terminate the legacy system.
- Have responsibilities been fully transferred over to the client/user?
By all means this is not a complete list or relevant to every project. Data migration requires a great deal of planning, experience, and expertise, some of which was covered here. For data migration solutions contact SDSol Technologies.
No comments:
Post a Comment