I recently accepted a Junior Systems Administrator position with the same company after about three years on the help desk. I also spent roughly a year serving as a team lead focused on reporting and weekly operational meetings, and I have around seven to eight years of total help desk experience.
As part of onboarding, I'll be completing AZ-900 first, followed by training on MECM and Windows Server 2022. For people who have made a similar move into systems administration, what do you wish you had known or done when you started? I'd especially appreciate advice about mindset, technical skills, documentation, mistakes to avoid, and making the transition successfully.
5 Answers
If you’re working in a Microsoft environment, prioritize PowerShell and Microsoft Graph. Automation becomes increasingly valuable as the environment grows, and some administrative functions are easier—or only possible—through command-line tools and APIs. A home lab or small virtual environment can also help you practice Windows Server, Azure concepts, and scripting without risking production.
Don’t let the certification requirements intimidate you. AZ-900 is mainly about understanding Azure concepts and recognizing the purpose of its services, so some focused study and hands-on exposure should make it manageable. Also remember that vendors may ask for excessive permissions or security exceptions. Ask why they need them, push back when appropriate, and don’t weaken your organization’s security just because a vendor requests it. Finally, stay respectful toward the help desk—you were in their position, and good relationships make troubleshooting much easier.
Document everything. Keep a searchable notebook with commands, troubleshooting steps, system details, decisions, and anything you may need six months from now. Write procedures for server builds, backups, upgrades, network recovery, and recurring tasks. Documentation helps you learn, makes communication easier, and gives the team something dependable when knowledge would otherwise be lost.
Congratulations! The biggest adjustment is learning to think beyond individual break-fix tickets. Start looking at the larger environment: what can be automated, upgraded, documented, secured, or improved before it becomes an outage? Learn how the systems depend on one another and try to understand why something works, not just which steps fix it. It’s completely normal not to know everything, but being able to research effectively and ask good questions is essential.
Be careful with production changes. Test in a lab or virtual machine when possible, make sure you have a rollback plan, and triple-check anything that could affect everyone. A backup only counts once you’ve successfully tested a restore. Understand your virtualization platform early, learn where the important backups are, and verify that fixes survive a reboot instead of assuming the problem is permanently solved.
That’s good to hear about AZ-900. I’m also looking forward to learning PowerShell and building a small lab so I can practice without affecting production.