In this post moving closer towards the end of my series on solution architecture for the Power Platform we’ll start to look at the considerations we should be making for planning a go-live and moving a solution into production and a project closure state.
Are you ready for last activities prior to go live?
So friends, if you’re in a state where you have loads of bug work items potentially still outstanding and features missing or UAT not completed, maybe you need to reconsider whether you’re ready for these last tasks and a production deployment to go-live! Be in a good project position and implementation position to proceed with these tasks.
Now let’s look at some considerations to fit into last parts of our engagement prior to being ‘ready‘ for a go-live.
Performance ready?
So friends, chances are in test and UAT we won’t have tested non-functional or ‘technical’ requirements. Its not something users will worry about so much but it’s still something we need to ensure we include in the process. One of the things to test here is performance for the solution!
Consider…
- Have you tested the data throughput you’re planning to chuck at various custom logic objects and your wider solution?
- Do you have a dedicated environment for performance testing with high volumes of data?
- Have you identified the key areas to test for performance to ensure risks are reduced?
- Do you have a plan (with automation potentially included) to populate the required data for performance testing.
It’s absolutely crucial that we perform performance testing friends. If it so happens that in production we end up throwing higher volumes at the solution than we tested and this hits limits, or becomes too much, this can cause serious service degradations for our users enforced by platform limits. Get this part right!
Deployment planning
So friends, the next consideration to make is a deployment plan. Now I’m talking less the technical ‘push to prod’ type thing here and more everything that happens around this. Here’s some thoughts…
- Production testing with stakeholders prepared
- Down periods
- Considering need for data migration
- Rollout strategy
- Support during deployment for any issues that arise, suitable for every severity
- Communication to ALL stakeholders – this is important!
Ensure that you have risk mitigation plans in place for deployments friends. The stuff above needs to be thought about.
Risk assessment
So people… are you ready for a deployment. We touched on this earlier in the post however it’s still important to assess various points even if we think at a high level that we’re ready. Consider…
- Where could things go wrong?
- What are weak points to ensure are tested post deployment?
- What happens if we have a production down issue?
- How easy is rollback or resolution for issues?