I've always believed that managing from afar doesn't quite cut it. Being a working manager can make you enjoy going to work instead of dreading it like a root canal appointment. It also helps you understand your team's challenges, and understanding is the first step to empathy. Maybe you'll even stop rolling your eyes when someone complains about a task you now know is a real pain in the butt. Here are 20 reasons managers of Linux administrators should be working managers, with real hands-on duties.

  1. Work alongside the team and show that you are willing to get your hands dirty. That example can inspire people to work harder.
  2. Stay close to the team's daily challenges, so your decisions about the work, the team's direction and its goals rest on firsthand knowledge.
  3. Know the system well enough to save it. HBO once came within four hours of going off the air. A talented developer had updated the code on both the primary and the backup servers that played video to air (a no-no), then left on a well-deserved vacation. The server was running, but it pointed to the wrong digital device, which would run out of content in four hours. Restoring from backup would take longer than that, and would also take HBO off the air. Because I had done kernel-level programming as a developer at Bell Labs, I could reverse-engineer the change, find the fault and switch the running code to the correct device driver. We already had a rule that the primary and backup are never updated together, and it hadn't held, so from then on the backup stayed offline and could only be updated by an administrator.
  4. Earn respect by being in the trenches with the team instead of spending all day in an office apart from it.
  5. Build stronger relationships by sharing the work. That shared effort can lift morale and teamwork.
  6. Step in when unexpected problems appear. Your team should know you have their backs.
  7. Do your homework before you call the vendor. Vendors often resist owning flaws in their own software, and they will have you spend endless time proving the problem isn't in your environment before they assign anyone to troubleshoot it. Because I knew the systems firsthand, I could cross those t's and dot those i's before the call and skip most of the back and forth.
  8. Take on-call shifts yourself, then engineer your way out of them. At first I covered about half of our on-call shifts. As the team wrote fixes for the recurring problems, I got that down to about 5 percent.
  9. Delegate with a clear view of each team member's strengths and abilities.
  10. Give useful feedback and performance evaluations based on the work you understand firsthand.
  11. Interview as someone who still does the work. You can't kid a kidder: you quickly spot the candidates who can't do the job at the level you need. And some candidates will teach you something, with a novel solution to an age-old problem.
  12. Spot risks and threats in the team's work, then take steps to reduce them.
  13. Prioritize with a clear view of the work. You know which ticket is urgent and which one is just loud.
  14. Build stronger client and customer relationships by understanding the technical work behind the service.
  15. Find places where the team's tools and resources can work better.
  16. Watch the workload closely so you can prevent burnout and support a healthy work-life balance.
  17. Know when to hand the keyboard back. A manager who takes every hard problem keeps the team from growing.
  18. Explain technical issues clearly and give targeted training and mentorship.
  19. Balance team needs with organizational needs so the work supports the company's mission and objectives.
  20. Stay involved in the technical work so you can see its effect and keep finding satisfaction in the job. So don't be afraid to roll up your sleeves and get in there with your team!