22
votes

OK, this is a non-tech question, but I think it pertains to a lot of sysadmins, especially at small shops.

There are a lot of tasks that are not related to a project, or day-to-day operations that do not add visible value to the organization, but nevertheless are important. I'm thinking of things like Documenting the Network or Processes, Researching that new tool/technology/process, development of subordinates, filing of paperwork.

All of these are important, but because they do not end up on a Gantt Chart, or directly result in your boss yelling at you if you skip it, can other be neglected.

How do you make sure that you cover these?

I have tried adding them to my task management system, scheduling recurring appointments in my calendar for things like "Research New Technologies". But these tend to be really easy to ignore/cross off this week, etc.

PS, I like the tag calendar-management, as the guys on Manager-Tools say, You can manage your calendar, but time is a constant, you cannot manage it to create more/less or to change how fast/slow it passes.

2
  • 1
    +1 for the Manager-Tools reference. They have a couple of calendar management podcasts which are enlightening and actionable.
    – tomjedrz
    Jun 13, 2009 at 18:21
  • Manager-Tools is a must listen for new managers, or managers who want to be more effective. Edited question to add link to Manager-Tools.com
    – BillN
    Jun 13, 2009 at 19:01

10 Answers 10

7
votes

Since this is subjective, I thought that I'd just toss out the ones that help me the most. All of `em can be found elsewhere, cribbed from "How to do better at time management!" articles, etc. They're just what help me. They may not necessarily help YOU.

i.e. Your Mileage May Vary.

Also, while they're not necessarily a direct and succinct answer to the question at hand, I believe that when used, they contribute overall to improving time-management, which has the direct overall effect of helping me free up some time, to complete those important-yet-not-urgent-tasks.

1.) Train everybody to do your job. Meaning; if you're on a team of 5, and there's one guy who knows MS Exchange, one guy who knows LAN, routing, etc. One girl who knows AD and core MS Administration and one guy who knows Linux/Unix, and you're the LAN/router dude.. Spread your knowledge around. Maybe the AD girl wants to know how spanning-tree protocol works. Perhaps the Exchange guy wants to know how his packets get from "a" to "b" to "c" and back. Work with them to get them trained up on some good basic knowledge. By doing this, you'll be more apt to go to them for help when you're inundated and can't seem to find the time to get the small-but-important stuff done such as documentation and that equipment refresh projection list to your boss that he needed, oh, say.. Yesterday. If you've done a good job training your peers, then you can ask for, and expect great help with relabeling those switchports or verifying SNMP on your devices and that they're all plugged into your monitoring system.

That being said, be equally prepared to learn some of their jobs as well, and for good reason. It'll help you understand where they're coming from and give them the latitude that you enjoy as well.

2.) Set an `email schedule' for yourself and limit your email checking to particular times of the day. Say four or so, if you work in a very, high-traffic, busy email environment. Focus on what directly applies to you and what requires immediate attention. STICK to the "check email at this time" schedule.

3.) Break tasks up into chunks of short-term time to be spent on them. I.e. multitask, but don't work too hard on a single task and get sucked into it, or burn out. Work on it for 20-30 minutes or so, then move onto something else. Keeping your interest piqued is critical to successful multitasking. it's altogether too easy for us to get sucked into troubleshooting WSUS and spending hours on chasing down one elusive hex-error code after another.

4.) Take a break every once in a while. Actually get up and walk around. Get the blood flowing. Go get some water. Clear your head, (what you were just directly focusing on) and think about something else. You'll be surprised how refreshing this is and how well and quickly it helps you focus on the next task at hand.

5.) Check out resources on the internet for Project Management. There's usually great tips and tricks in there that directly deal with questions like yours, along with general time-management skills. You'd be surprised.

6.) This is gonna sound weird, but work at being lazy. Learn how to automate, automate, automate. Don't waste your time repeating tasks over and over. Learn how to do it with a script if you can. I suppose that the proper way to say this would be, "work smarter, not harder".

7.) Use a SIMPLE to-do list broken up into daily, weekly and monthly. Prioritize the tasks in a similar way by, "critical", "not-critical" and "long-term". STICK to the list. Update it, don't let it get stale. Make it the first thing you do every morning and the last thing you update before you leave. This does wonders for continuity of thought, etc.

8.) If you work in a meeting-heavy organization and you find your time being taken up by tons upon tons of pointless meetings each week, use this philosophy; If invited to a meeting, does it directly apply to you, or are you a key stakeholder? If not, consider how your time could be better spent. Especially if you get an emailed summary of the meeting minutes. Meeting invites usually (in email) come with, `Accept, Tentative, Decline' options. Don't be afraid, you can use them.

When attending a meeting, if there's no clear agenda or one displayed in the first five minutes, leave the meeting immediately. Get out. 9 times out of ten, it'll ramble on for an hour or so, with no point. It's the equivalent of sending a team of people to the grocery store with no list. What they come out with will be a hodgepodge mix of foodstuffs. Much like what you'll walk out of that meeting with.

Anyway, again, check out the PM references out there on the internets. They're worth it.

HTH..

1
  • "This is gonna sound weird, but work at being lazy" - Totally agree with this. It sounds counter-intuitive, but most of the 'good admin' skills I know are emergent from underlying laziness. Why repeat things if we don't have to? We want to do the bare minimum of the repetitive and boring stuff, so we can focus on the different, interesting things. "Work smart, and hard". Sep 12, 2010 at 21:53
8
votes

Getting Things Done is a great book, very helpful, even if you don't implement a full system. GTD is a very long commitment, don't get overwhelmed.

Here's a brief intro to GTD.

But no system is perfect. It's very easy to lose sight of WHY you are trying to get organized, if your organization system takes over all your free time.

The map is not the territory.

For a more domain specific book check out: Time Management for System Administrators by Limoncelli. Which will provide a few more insightful tips.

I personally also use todo.sh. Which is a great todo bash script. It's also the only bash script I know that's unit tested and under revision control.

2
  • 1
    +1 for mentioning GTD and a new O'Reilly book that I'm gonna get now, thanks! :) Jun 15, 2009 at 19:45
  • +1 for mentoring todo.sh
    – adopilot
    Feb 20, 2010 at 16:07
7
votes

I found that the best way to take care of these tasks is to come up with a scheduling system, so that A) you are not spending all of your time on the day-to-day problems that come up and B) you reserve time for the documentation and other back-office issues that need to be taken care of.

In that light, I would highly recommend David Allen's Getting Things Done (GTD). His system of managing ever changing TODO lists fits in well with how administrators and managers work. I've been using it for 3 years now and my time is much more productive, and I never feel like I don't have a handle on all of those little issues that should be handled.

I'm sorry if this feels like an infomercial, but his book really is that helpful.

5
votes

I like the "handle it only once" technique: whenever possible, get rid of things on your to-do list as soon as possible.

  • When you some form from HR in your inbox, do it immediately and send it back.
  • If you fill out timesheets in your company, once you start on it don't stop until it's done.
  • For documenting processes, the next time it's a bit slower and you're doing the process, type up the process as you go (or update the document if that's what needs doing).

For other documentation tasks, even though they add a bit of time to the job, you have to get into a habit of doing them. e.g. if you have a spreadsheet of switch/patch-panel connections, you have to take the time to update it whenever you make a change, otherwise you'll spend more time later sorting things out. Similarly, we keep our patch panels very tidy with lots of velcro tape and it takes a little bit longer to re-do all the velcro when we change patch connection, but it's worth it because you can see where everything is.

1
  • I'm a fan of properly redoing the patch panel charts monthly, but having a "quickly write down what you change the minute you changed it" spreadsheet taped right next to the real chart.
    – phuzion
    Jun 16, 2009 at 14:23
4
votes

How about replacing "Research New Technologies" with more specific items, e.g.

Spend 15 (or 30) minutes researching SELinux and work towards answering one of:

  • How do I install it?
  • What mode of operation best suits my environment and how do I configure it?
  • How do I interpret its log entries?
  • How do I create rules?

Similarly for "Documenting the Network" etc.

4
votes

I like "personal organization" as the moniker for this kind of stuff.

There are two separate things here. The first is getting a system in place and being disciplined enough to follow it. There are a ton of systems out there .. GTD gets lots of press. Personally, I use a hybrid of Franklin and GTD. Have a framework to capture the things you need to do and to let you easily prioritize and schedule them. I suspect that this is where you are breaking down .. as you noted, you scheduling then crossing out.

The second is to have a way within that framework that Quadrant 2 stuff {important but not urgent} actually gets some attention. Quadrant 2 is the place of real, lasting value to the organization (and to you personally), and usually does not get nearly enough attention.

You can start by not thinking of (or referring to) them as "me" tasks .. if they aren't important to the organization you shouldn't do them on company time. If they are, they need to be folded into the mix somehow. If they truly are about you .. apply the same principles on your own time.

The best way I have found is to treat these things as any other project. Any good personal organization system has a process to capture and include the tasks from projects. Take the amorphous ideal (say, Document Server Configuration) and break it down into the smallest tasks (Build empty doc, doc network config, doc shares, etc.) you can get down to. Then include some of those tasks in your daily/weekly plan, and find a way each week, to get those little pieces done. Note that I said each week, not every week. It is kind of like recovery .. today is all that matters, you work your system today, and tomorrow you worry about tomorrow.

This won't get you all the way there, but it is a start. In the end, this is a relatively small extension of the discipline to keep on a system. But it is the part that turns you from "getting the job done" to "making things better".


EDIT - respond to comment from original author.

Be clear - this discussion is NOT about those items that are "low organizational value". You should not waste time on unimportant tasks unless you have nothing else to do. This dilemma (and my answer) is about tasks that are HIGH organizational value, but not time-sensitive.

One additional strategy, that I recalled thinking about the comment. I have found that virtually all Quadrant 3 (high urgency, low importance) tasks can simply be ignored or deferred until convenient. The trick is knowing which urgent items are important and which aren't.

2
  • I've never heard the term "Quadrant 2" until now, but I've used the concept: some things are important but not urgent. I also try to spend as little time as possible on the other combination: urgent but not important. Jun 13, 2009 at 17:52
  • I didn't like the Title "me-tasks" I started with "low organizational value" but that wasn't right. I tend to think about these as important "to me" but the users(the organization) could care less. Important but Not Urgent is the best description.
    – BillN
    Jun 13, 2009 at 19:12
3
votes

I think that is one very important thing to fight with taht, Is that you do a good segmentation of tasks,

Draw a circle divided in four fields, for each field in circle give atribute

  1. VI- LU Very important less urgent
  2. VI - VU Very important very urgent
  3. LI - VU Less important very urgent
  4. LI - LU Less important less urgent

Then You can in nice way segment jobs and , do cross working for better picture where is time gone. Some time I said to myself for two of VI LU task You can pick up one LI - LU.

In general they said that position one in my list should be addressed to investors, owners, and people who making big decisions but in no hurry,

2nd step is for Top managers whose job is fireplug

3th and 4th are for lower ratings of employers

2
votes

I find I end up learning new technologies on "my time" and I view it as "continuing education" and something good for my career.

1
  • Agreed, I think we all do this. However I expect that most also put in long hours, and need to find a balance between work and home/family.
    – BillN
    Jun 13, 2009 at 16:51
2
votes

Make doing these things part of the task that generated them.

Installed new software? Include documenting it in part of the plan to install it.

Staff Development - this can be an ongoing thing throughout all new tasks aside from the one or two specific training things, and these are important and hence deserve to be given the time they need.

Paperwork to do? Include that in part of the plan for the task that generated the paperwork.

Block out a certain amount of time for administrivia each day - if you don't let a backlog develop then keeping up with it will be easy.

No time for research? This is also an important part of the job, not a trivial addition. If you don't have time for it then you're either under-resourced or you have another problem if you're constantly being dragged away from stuff like this to fight fires.

1
  • +1 good idea. A nice side-effect is that it shows the customer/boss/whoever gave you the task the true time cost of doing something.
    – sleske
    Feb 17, 2010 at 10:50
1
vote

Probably not the answer you want to hear but I've found that I get more time for that kind of work the older I got - people just give you more time and don't 'rag' on you so much.

If you show you're always happy to take crap from people then they'll carry on doing it, once you're older you get less people trying this so you can pick and choose how best to benefit your organisation.

And yes I know we're the same age :(

1
  • 2
    Chopper3, I know, I find that as I move up the chain, it is easier to do this. The question was as much for newer sysadmins as myself.
    – BillN
    Jun 13, 2009 at 19:17

Not the answer you're looking for? Browse other questions tagged .