Slashdot Mirror
Realistic Sysadmin Workload for a Company of 30?
An anonymous reader asks: "My company was recently sold to a new owner. Currently I am working as a programmer using a number of languages (Java, C, C#, PHP). I am the only maintainer/developer on a number of important code bases. The new owner wants to add 'Network Administration' to my list of responsibilities. We are moving locations and our infrastructure needs to be rebuilt from scratch. He claims that after being set up (something I am also responsible for) our company IT needs can be met using only 1% of my work week. Our user base will be 30 people, mostly programmers, with a minimum of non-techie staff. I am a professional programmer, but have no real sysadmin/network admin experience. His solution is 'We'll get you a book'. Learning new things is great but, I just want to be a programmer. I'm worried that this network admin responsibility will become my new full time job. Does this 1% statistic hold water?"
I know a guy who was the primary programmer at a similarly sized company and also the lone admin. He consistently worked weeks of over 40 hours. Since programming was his first priority he rarely did admin stuff. Low priority admin tasks would never get done unless the projects really really dried up. High priority admin tasks would mean overnights and terrible times.
The boss likely doesn't want to hire a separate admin since that person doesn't make direct money for the company. A programmer makes software which brings revenue. An admin makes computers work, but doesn't bring in any direct revenue.
If you are moving there will be a lot of up front admin work. If you can set something up that is really kickass from the get go, then you can probably keep the amount of admin time per week in the future really low, but not down to 1%. Of course, this requires basically not programming for awhile just to plan and set everything up. But if you don't then the admin work will be this ghost constantly haunting your higher priority programming.
The GeekNights podcast is going strong. Listen!
Whilst i'm not convinced about the 1% value, It is possible that that might work in a correctly, carefully set up network environment where each users accesses & rights is carefully set up, and you have a hardware support contract with someone, but I doubt it ;) ), plus i'll wager that those 30 odd people will mostly be running windows, and will have local admin rights - that really increases the difficulting in managing them, especially if they are connected to the internet in some way.
However, irl this is *not* going to happen.
for a start, you are not going to be able to plan and set it up right first time (thats where the experience bit comes in
Basically, your boss is being a cheapskate. You *need* a sysadmin, or at least someone whose job is officially part sysadmin and has experience - ask the boss whether he would want a sysadmin with little no programming experience and 'a book' to be writing the core code for your product? I suspect not. So why does he think the reverse is true?
"Success is based on knowing how far to go in going too far"
I mean, sure, once the network is set up, the infrastructure for 30 machines should be perfectly stable.
But then email stops working. Or someone gets spyware on their machine. Or a graphics card plays up. Or someone loses their printer settings. Or a mouse is playing up. Or someone can't get through to google.
As Sysadmin, whenever anything goes wrong you're the person they'll come to. If you're working purely with techies who can handle most problems themselves, then fine. But if there are _any_ non-technical people in your company then I'd estimate 25% of your time will be spent dealing with them.
However, your boss isn't going to listen to this. So what you do is find a free help-desk package (if you're using Windows then Liberum is pretty good) and get people to funnel all of their support calls through that. That way at the end of the month you can go to your boss and say "Look, this is the amount of work it takes to keep a network up and running. That's why I haven't got any programming done."
My Journal
Talk to your boss about security and tell him that it's a process not an investment and need a steady (time) budget.
Would be interessting what your boss answers.
Grundgesetz * 23. Mai 1949 - 30. November 2007 - http://www.vorratsdatenspeicherung.de/
That most executives with this kind of understanding of IT infrastructure (i.e. little to none) tend to confuse systems administration with tech support. Sounds like you're being asked to fill more than one set of shoes.
As a professional systems administrator myself, I can tell you that very few individuals posses the capability to both program and maintain a mixed network. I'm not saying it can't be done, but it usually requires giving up more than just one's wishes to stay in their area of expertise. It also requires giving up weekends and vacations, as you'll inadvertently become married to the machines as more time goes by. It's unfortunate that IT professionals have gone from being held in high esteem to the average corporate foot soldier, thrown about at the whims of unknowledgable people, and ultimately, expendable. Good luck with your situation.
End of Line.
If you don't know, how would he know what he's talking about?
I do both, and let me tell you it's more like 30% than 1% - and I'm not even doing everything. Not that it's not enjoyable, but proper sysadmining is a really important job, it's making sure everyone else is working smoothly. If it's badly done, the productivity of all these 30 employees will be affected.
If this is the case and you doing system adminstration for 30 people will only take 1% of your time, then the sysadmin work load / person is around 0.0003. This would mean that a company in a similar industry with a staff of 100,000 employees would only need a sysadmin crew of 30 people. When you think of it in those numbers, it immediately becomes apparent that the numbers are not even close.
From another angle, I would ask your boss why he has an admin, a marking/sales person, and/or an accounting person. The accounting work for a 30 person company has to be only a 1% work load for him. He can do all the administrative work in 1% of time. And there is absolutely no reason he can't take care of the sales and marketing items in another 1%. That's only 97% of time. What's he going to do with all that 97%?
As has been said before, there are real professionals who do systems administration. There are some people who can do reasonably well at sysadmin, network admin, network design, systems design, programming, etc. They are rather rare and they can't do all of them at the same time. For a company your size, it would probably make sense to get a person who specializes in sysadmin and can program a little bit (understands the code enough to be able to read and possible fix some stuff) and the two of you would work as backups to each other.
It's not clear whether you're expected to be the systems administrator, the network engineer, or the all-purpose all-singing all-dancing IT guy. Let's examine all three scenarios.
We'll suppose you work a 50 hour week. 1% of that is 30 minutes. In the "network engineer" circumstance, that's about enough time -- assuming that you have a very well designed and stable, simple network built on the most reliable hardware available, and you never change anything, just fix it. That won't happen, of course, because you've never done this before and therefore you won't get it exactly right the first time. I won't even mention that your boss is a cheapskate who won't be buying the most reliable hardware anyway. The first time you need to deal with your upstream ISP will chew up 30 minutes. If you ever need to buy replacement hardware, that will take a few weeks' time as well.
Now, as a systems administrator for 30 people, plus maybe five or six servers, you'll blow through your 30 minutes of allotted time every Monday before lunch. Someone needs a password changed. Someone else says "mail isn't working". The sales critter hands you a laptop and says "I spilled beer on it, can you get my files back?" Those are just the incidental time-users. When are you going to upgrade your antispam system? There's an intermittent problem with one of the file servers. Diagnosis may take more than half an hour.
Do I really have to say anything about being the defacto IT shop? No, I didn't think so.
Tell your boss that you want to keep track of your IT hours and be paid for everything over 45 minutes a week at the same rate he would pay an outside contractor. Since he's certain that you'll never go over 30 minutes, this is a great bet for him.
You should start looking for a new job with management that can make more realistic predictions about workloads. Meanwhile, explain to your boss that you heard that your coworker runs a network at home -- maybe he's a better choice?
Are you going to learn how to be a sysadmin and network admin on the clock? Reading a book won't be enough. You'll need plenty of time, especially if you want to effectively secure your hosts and your network. My guess is he's not willing to pay for your time, especially not while your projects stall in the meantime.
There are consultants that just do setups. If he wants it done right, but is too cheap to hire a full or part time guy for just the servers and network, he needs to look at this as the next-best solution. At least, if they screw up, they can be held responsible. And then, as needed, either you or someone else can make minor modifications as situations warrant. Do you want to get blamed if the book you got and the weekend of cramming wasn't thorough enough to stop a scriptkiddie from 0wning j00r cvs server and erasing it, or worse, a competitor rootkitting it and installing a backdoor so they can watch your progress, maybe change some data, a couple months down the road while you're too busy on a real project to track vulnerabilities and new attack types in the 24 minutes a week allotted to this? (less than 5 minutes a day... can you even get through your email that quickly?)
Oh, and I'd say, get your resume ready. If he starts having more unrealistic expectations of his staff, you should probably look to go elsewhere.
You might think that programmers are easy to cater for as a sysadmin but you probably couldn't be further from the truth; programmers and other tech-savvy people will install programs, change OS settings, (un)plug cables, change BIOS configurations or whatever they have access rights for (if not; they might try to hack the OS to get these rights). It's a lot easier to support people who just use their computers to read some mail.
Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
If you do buckle under and play it his way I can guarantee you within a year of moving to your new office he will be blaming you for "not reading the book" for every extra minute you spend doing sysadmin work-- Likewise, you'll be blaming him for pushing you away from programming your programming career by insisting you "get the job done right first" with your admin duties.
Take a stand if you wish, but most small businessmen operate on the principle "No-one is irreplacable" and that means you too. You'd be alot happier working for someone who understands different IT roles and understands what your personal carreer goals are.
Height: 38U, Weight: 0 Newtons, Eyes: #0000FF, OS: Gray Matter 1.0 (Alpha)
MOD PARENT UP! Very true, but a little too mild, in my opinion.
The job that is mentioned in the Slashdot story would take an already skilled person 50% to 100% of his time. That's because it is not serving regular users, it is serving programmers, who expect a lot more from their computers.
Computer administration is not just administration. There a many lengthy one-time projects, like finding better backup methods, or dealing with the latest vulnerability. Fixing and cleaning after a serious security breach can take a month, for example.
Anyone administering Windows computers must deal with the fact that there are people with huge amounts of money who want to exploit Microsoft's (deliberate) sloppiness. One list of major investors in spyware companies shows a total of over $139 million in venture capital. Remember, Microsoft makes more money if a user becomes tired of slowness and problems caused by spyware and buys a new computer, which is how most resolve such problems. If you administer Windows computers you have the richest man in the world and his rich think-alikes riding on your back.
It sounds like the old story. People with control over more money than brains buy a successful software company, figuring that they can extract more that ever before from the customers.
We already have enough information to predict that the company will go out of business. Because it is a reasonable assumption that the person who submitted the Slashdot story isn't the only one being abused, we know that the company has already begun dying; the abuse is killing the company right now. It may, however, be a slow death, sometimes old customers are reluctant to change to new software, and try to live with the new stupidity.
There is a reason why Dilbert is one of the most popular comics in the United States. The real bosses are actually worse than the pointy-haired bosses in the comic. The real PHB's abuse everyone, take more than their share of the money, and destroy the company, too.
The new owner of the company is wanting to test the limits to see how much he can abuse the Slashdot story writer. He is: 1) wildly out of touch, 2) ignorant, 3) self-destructive, 4) arrogant, 5) abusive, 6) seriously abusive, and 7) lacking in social skills.
What may happen is that not enough time will be spent on computer system administration, and the programmers will not be served. That's the self-destructive element.
Not to mention these fun possibilities: ...) and your boss who doesn't value the admin side gives you bad reviews on your performance evals.
You get behind on your programming projects because you're doing admin tasks (backups, patches, testing patches/backups, checking logs,
You become the scapegoat for everything related to system failure. Hardware fails, you didn't do your job. Software patch creates unexpected software failures (this happens more frequently if you use 3rd party tools in addition to MS products), and any other thing that might fall at your feet.
There are many more examples but if your boss/company doesn't value the job of sysadmin you're not going to get any points for the work you do. Especially since he thinks it's 1% of your time. What a crock of shit.
I handled it by just ignoring the dumbfuck boss and finding another job (which is working for myself so yet again I'm the programmer and sysadmin, at least I respect my own work!).
I was in a similar situation a few years ago. 45ish people, we rolled our own network, mostly techie types...
:-/
We needed about 1.5 system administrators.
Fortunately, we had two. So about 1/4 of my average work week was spent as a testing droid for the developers and-- get this-- getting ahead of the game.
Whoever told you 1% of your work week is on crack. Stuff simply just doesn't work that well.
-JDF
If there is one thing you take away from this post it should be this: the way people who favor a course of action process information is different from the way that people who disfavor it do. People who favor a course of action are very clear on the intended consequences, but only vaguely grasp it's unintended consequences. People who are against do the opposite.
When the people above favor something and the people below disfavor it, you get the "what were they thinking?" anti-pattern. When the opposite is true, you get the "dinosaur business" anti-pattern. In most businesses I have seen, you have on one hand people who are frustrated because people are too pig headed to change, and on the other people who are feeling pestered drop the work that has to be done in favor of some hair-brained scheme. Your job is to get this out of the realm of impulse and passive aggression into the realm of rational decision making.
There are a couple of strategies you can use here.
I can tell you often the quickest and easiest way to forestall a bad idea is to accept it's presumed benefits as given and cheerfully take on the job of planning to make everything work acceptably. You just need management to decide between some options you've come up with to handle some implementation details. Not passive aggressively chosen options either -- the best ones you can come up with. For example, if you are on vacation, even if you have a beeper, you won't be able to fix the email server. A part time admin like you could probably get control of the worst situations in, say two days if you're on site, but unless you have a second developer comparably involved it might be as long as a week. So, you get an estimate from an IT services company of what it would cost to have somebody come in on an emergency basis. See -- an undeniable problem scenario, and three options: double the effort, hire a consultant, or accept that there is a possibility that email may go down for up to a week. Continue cheerfully running down the list of dealing with all the problems that are undeniably possible, and all your reasonable solutions to these problems, until their resolve crumbles. If it doesn't make sure you have their commitment to each of your solutions, or to accepting the responsiblity for the risks involved.
This is a good strategy to take if you think that management commitment to this idea is shallow. More often than not people are looking for a quick fix, and enthusiasm evaporates once things don't look so quick.
A second strategy is to actively and frankly sell the idea of a professional adminstrator. Right off the bat, I'd say "I understand the benefits of this company of controlling overhead costs, and that a network administrator is a significant expense whose benefits are hard to measure on the bottom line. But I'd like a chance to show you that a professional administrator would be more cost effective." Then you ask to have a chance to do a little research and put together an analysis of the alternatives, which of course he'll understand is a sales pitch. This is a mode of decision making that managers understand and respect.
This is the most generally useful approach, but it depends on your salesmanship. You need three things: (1) knowledge of what would make the customer buy your product, (2) understanding of the ways the customer likes to make decisions (3) the customer's trust. You have to prepare your analysis of the customer and the pitch; try to find out what his hot buttons are and make sure you push them when the time comes to close the deal.
The outline I think is pretty clear. You examine why the best run companies in this industry use pros to administer (you don't have to establish this is so, or anything else that sounds reasonable). You show how network outages would have interfered with something that was important to the boss, like getting the proposal on the big contract out on time, and assign a round percentage chance of say 1%. You multiply this by the to
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
I agree.
Nobody here seems to have done the numbers, though. (Surprising for a geek site!)
1% of 40 hours is 24 minutes.
Make a list of activities he wants done and activities that are done weekly in a normal sysadmin job. Estimate the times it takes a trained sysadmin to do them and add up all the times, then point out that as a new sysadmin, it'll take 5-10 times as long on many tasks to learn them, and a few months before you can get the timing down to only a short time per task.
Present real numbers to him so he can see that it takes more than what he thinks. As someone else said, a sysadmin does not directly show up in a profit statement (a sysadmin only enables others to do their job with their computers without having to think about it) and the new boss is looking at the figures. If he hires a sysadmin, that's a lot of money he can't keep himself or use for something else, so he will not want to hire one unless he has to, or will try something like getting a part time sysadmin.
Also, once you get started on sysadmin work, start logging your time. Make sure when you talk with him, you aren't just saying, "System work took up 5 hours yesterday and 6 hours today," show him, with a log what took up that time and why it was important. That way he can't say, "Cut the time down," since you can show that you spent that 6 hours today doing things that had to be done. If you list 15-20 tasks that took 6 hours, that is way better than letting him think you were just slacking off and taking your time.
I have to agree on the "I just want to be a programmer" part, but perhaps with a different spin.
See, computer programming is different from system administration, just like being the CEO of a company is different from being the CFO, or being an engine rebuilder is different from being a transmisison rebuilder. Trouble is, most non-techies don't realize that, because they have no idea what techie people do.
So I suspect this person's new boss used to have a guy that did -everything-, and possibly did it well. That guy (or girl) worked long hours, was underpaid, and eventually got burned out and bailed. So this new boss probably thinks that all techies can do -everything-, and just assumes the previous person's poor attitude was responsible for their burnout/departure. This is an opportunity to teach the new boss that not all techie jobs are created equal, and not all skillsets transfer over.
On the other hand, saying "I only want to be a programmer" will be interpreted by his boss as "I am comfortable where I am and don't want to grow". This may be unfair, but that's how it will be viewed, and that's a bad thing.
Ideally, what this person might do is talk with the boss, and explain that programmers and system administrators have two very different skillsets, even though they have similar technical aptitudes, just like carpenters and electricians do. Explain that you'd love the opportunity to learn that new skillset, but it's going to take more than just leafing through a book, because keeping a network of 30 machines alive is a full-time job even for an experienced person.
Further, explain that it doesn't seem like a full-time job from the outside because the work comes in fits and spurts, where one day you might do very little, but the next you might have to work overnight to get things fixed, and most people outside of system administration have no idea those overnights are happening, because they're at home; all they see is an idle system admin sitting at a desk on the good days. Oh, and mention that you know all this from talking to a few system administrators that you know.
Finally, tell him that you will be happy to take it on, but it won't be practical unless the following conditions can be met:
1. You will have to take formal classes to learn how to do it right, at the company's expense and during paid work hours, so that you can do it efficiently and quickly when trouble arises;
2. When trouble arises, programming projects are going to be put on hold until the trouble is solved, and so programming deadlines will always need to slip by the number of hours or days it takes to solve the problem -- and those slipped deadlines are going to cost the company money;
3. Even when there is no obvious trouble, a certain amount of time must be put aside each day to do routine maintenance and take care of users' workstation issues, because jumping back and forth between administration and programming tasks will make any person in that role painfully inefficient;
4. There will be times that system administration tasks require late nights, overnights and weekend work, and it is only reasonable to be able to get comp time (or overtime, depending on if you're salaried or not) for those hours.
Will the boss like this? Probably not, but you're not saying "I won't do it" -- you're giving him/her an honest and intelligent assessment of the situation based on your own research into the problem, and you're giving him/her a plan under which you CAN take on the role. Of course, chances are the new boss will find the plan to be less than ideal, at which point you might suggest a part-time contractor system admin or whatnot.
And of course there's always the chance they will say "fine, go do it". If that happens, and they hold up their end of the bargain, then congratulations -- you've just gotten paid to learn a valuable and marketable new skill. On the other hand, if they don't hold up their end of the bargain (claiming you never talked about that, or "I misunderstood you -- well, just get it done for now and we'll worry about your hours later" and so forth), you have to acknowledge you're working for a sociopath, and you should leave.
Good luck to ya, buddy.