Information Technology Lifecycle Management for SMB
A good friend of mine NickMac from NZ blogged about this a while a go and I thought I would do my own post talking about some of the applicability to this type of management to the SMB.
The slide below comes from Microsoft's own IT team. Now this is clearly designed for a large org but I think there are some great pointers to take when you are managing smaller IT as well.
- Have a plan - know what projects are coming up, think about how they will have an impact on other long term plans. Having a plan means you can prioritize and it means you can track your progress. One of the better tips I got when it came to planning was to always keep abreast of emerging technology, and always look on how to maintain your existing. My rule of thumb is once I choose a technology platform, track maintenance for 2-3 years, then track emerging changes from 2.5 onwards. This way you get the most out of your investment and then as you look to change you have plenty of time to research the emerging technology. Remember newer my not always mean better for your solution, and keeping something that's old may also impact your environment. Tracking these things helps keep you environment nimble. Finally always, always align your IT plan to your orgs business plan.
- Delivery - so you have your all up plan and you've made the decision to implement some new technology, now you have to plan the delivery. Delivery is made up of lots of things, first up, understand the task at hand, who it affects, why it affect them, understand the systems impacted, and have a tested backup plan. Taking the time to plan delivery and documenting this always makes the process much less stressful and more successful. You notice I called out some human elements to planning too. Understanding and satisfying your internal customers is the key to any successful IT project.
- Design and Rehearse - make sure you practice the installation and setup per your plan, with virtualization there are no excuses anymore for not doing this pre-deployment practice. Take lots of notes and remember the effect is has on certain applications. In SMB we're not going to deploying each solution too many times (remember Enterprise has a single person specializing so they do the same deployment time and time again) so doing it at least once before means you'll be more confident and more likely to get it right. Test a recovery too :)
- Backup and recovery - I know you would be testing this all the time but I'll put it in just to be safe.
- Do it - Ok now the main event, take all of the learning's you've done to get here and deploy methodically, if something goes not as expected, review your notes and make a calm decision. One of the common mistakes I see here is to start patching or move out of plan and find yourself in rat-hole of fixing issues caused by your fixes. Losing a critical path is cataclysmic for any IT project. Once you're complete move to the next stage
- Testing - you should have a sound testing plan from point 2, test, test and test again.
- Rollout - go live, and watch how everything performs, the first 48 hours are the most critical, keep testing as above and ensure that all system respond as expected.
- Close - right if you've followed all of the steps above you should have a big pile of documentation. I really like the Wiki feature in WSS3.0, set up a site for the project and put all of the information you've just created into it.
- Maintain - If you make changes update your wiki.