Slashdot Mirror
How Many Admins Per User/Computer Have You Seen?
miffo.swe writes "I'm trying to find the normal ratio of technicians/support tech per user or computer in your average IT-shop. When searching around, I can't find that many examples or any statistics. We manage around 900 computers (mostly Windows XP) and 25+ servers (mostly Linux). There are around 2600 users of varying knowledge, mostly pretty low. I can't find any statistics on this, so real-world examples are very welcome since we do this with one sysadmin (me) and two sneaker techs. Are we seriously understaffed, or is this normal?"
Over 9000
If you properly plan your implementations, there shouldn't be a need for many admins..
The real question is are you always constantly working your ass off, fixing stupid problems - and therefore unable to do anything more productive? If so, then it seems you don't have enough people.
If you have a fully managed office, and you can remote in to all these desktops and fix everything really quickly - then you're probably OK.
Like most of IT, whatever works.
I have developed a truly marvelous proof of this comment, which this signature is too narrow to contain.
It sounds like you are on top of things, but I'd say you are understaffed. We have 10-15 Windows servers, and a few hundred XP/Vista machines, and one admin-type, plus two full-time equivalent techs, and are looking for at least one more.
LRN 2 SWM
It's normal. With the economy and the dwindling amount of resources being set aside to IT in companies, you can expect to see the tech to end-user ratio reach 1:1000 often.
we're at around 1200 users and around 8 help desk people to support them all. 2 DBA's for 30 some MS SQL servers and 3-5 admins for 200 some windows/^nix servers. some people double by helping users in their office
At my company, about 1200 employees, a vast number of servers (mostly virtual). 4 service desk techs, 4 network operations personnel, 4 programmers. We still consistently feel a bit understaffed, but we manage. There are other locations of the same company with similar ratios.
Here's what we've had at different jobs:
Internal Corporate Helpdesk - 6800 users, supporting every application on desktops, 10 support techs during the day, 1 on nights and weekends.
Website support: 10,000 users, supporting general usage of just 1 website. 4 techs, regular business hours only.
You never expect irony, do you?
Want to be a professional wrestler? Visit www.iyfwrestling.com
@iyfwrestling
I remember hearing something along the order of 200 users, 50 computers and 2-3 servers per head in the IT department would be optimal when I worked as a sysadmin 10 years back.
I can tell you this much, administering 250 users, 120 computers and 7 servers is too much for one person atleast, that's from personal experience. =) Mostly Win NT 4.0 back then.
Right now its 4 offices around 120 employees and just me...oh and I forgot (or selectively blocked) a former client who keeps calling me to pick up after their new "IT guy" who is supposed to save them money. If they were all in one location I could probably juggle it better but as it is I'm starting to burn out.
This is party due to our lack of automation - Active Directory's not got much penetration outside our area, we haven't got automatic package rollouts/updates, no out-of-band management, and there's no planning WRT buying computers; each dept will buy a machine as funds & needs dictate, with input from us.
The three of us are desktop support. That doesn't count the sysadmins & netadmins.
Hail Eris, full of mischief...
E pluribus sanguinem
I've seen one sysadmin per 70 Unix servers and one sysadmin per 30 Windows servers. That's a general guideline for SERVER systems. Desktops are another matter. I've yet to see a serious roll out of Unix desktops, so I'm going by Windows systems, but one help desk tech per 50 systems is what I've personally seen as optimal. More Windows PCs per tech and the help desk gets overwhelmed. Less than that and they sit around and play games most of the day. This is assuming that you push updates over the network, not go around and manually update each PC.
From 2004 to 2007, I was the sysadmin for a high school:
- 2 full-time IT staff
- 3 WS2003 servers
- 450 computers
- 100 network printers
- ~80 video projectors
- 150 employees of varying knowledge, mostly pretty low
- 1800 students of varying knowledge, mostly pretty low
- 170 phones. PBX, not IP.
This seemingly short staffing was offset by:
- Using Altiris to centrally manage imaging and software deployment; a practical godsend.
- Ditto for being on a domain: Group Policies are very powerful if used properly. Some schools are just on workgroups.
- A well-run school district IT office, so I wasn't flying solo. They configured all my switches and WAPs, did server names and IPs, assigned me a DHCP range, and other "big picture" stuff.
In my experience, it's a function of how well you're doing your job and how locked down you can make the users' systems. If you do your job well and can effectively totally lock down the system (users install nothing, use Citrix, etc, and are only allowed to use a limited number of apps), that can be perfectly reasonable.
Government facility:
3000+ PCs
2600+ users(yeah I know we have more PCs than users)
200+ servers
6 Server Admins (understaffed)
2 Network Admins
2 Telecom Admins
3 Infrastructure techs
15 Helpdesk Technicians (overstaffed by about 5)
47 other IT employees for software support/dev staff and management staff
It does not seem widely off - I used to manage a team of 10 people who did the admin/networking for 8000 users. We did not do all of the end user hand holding, though some days it seemed like we did. Our ratio was about 1:1 users to computers, so your needs may be higher with the larger percentage of users. Rules of thumb are useful, but in any support function is it really important to keep data on what you are doing. If the team works really hard it seems like you aren't even there. Most people don't understand the "magic" required to keep this stuff working. If you track calls/time, it is much easier to ask for more resources/staff when they are needed.
We have a small shop, we support around 150 users, all on XP boxes, 2 Windows Servers, and 2 Linux Servers, we have 3 of us in our shop including the IT Manager.
I am Bennett Haselton! I am Bennett Haselton!
It usually varies every place I've been between the quality/age of the hardware and the competency of the users. Additionally it depends on how automated the system is, and whether there's a dedicated support staff. Small places I've been I've find you can do about 45-75 comfortably... It was a bit stretched when it reached 100:1
Just my $0.02
------
"And may your days be long upon the earth."
we have (at a small hosting company) about 200 freebsd servers and another 200 freebsd virtual machines (spread over 8 vmware-esx machines) which we manage quite easily with the three of us.
You have more than double that so i'd say you are pretty understaffed. I saw a video once that was actually pretty intelligent in talking about standard support ratios. Basically, there isn't a "standard" the answer is almost always "it depends". You start with your userbase - how tech savvy are they? How many applications are you supporting? What kind of hardware do you have? How many remote supporting tools do you have to use? Each of these answers adjusts the support ratio up and down and sometimes something as low as 75:1 is needed and other times 300:1 is just fine.
Still, in the place I work now we have 600 machines and 40 servers or so (most virtualized) and we have 13 IT people (with 1 open position right now). This includes 1 helpdesk person, 2 programmers, 2 systems support personnel (they support specific software we use), 2 hardware techs, 2 network analysts, 3 systems engineers, a secretary, and the boss.
"To strive, to seek, to find, and not to yield." - Tennyson
The contract I work on has one admin per 75 seat requirement. you need to get more people now!
Google doesn't have any relevant hits for this phrase (except this article).
both: yes
Due some reasearch on this and you should be able to get easy results...
I think the current ratio is 1 IT person per 100-150 computers, but it depends on what your users are doing.
Unfortunately you are forgetting some major details. Most importantly what applications are the users mainly using? For example their business system, is it proprietary? Do they seek help internally for financial problems within their financial system? I've worked at places where IT was nothing more then the geek squad to where I am now which IT seems to be a core pillar of everything. I have learned more accounting over the last 4 years then I'd ever wanted to know.
Additionally, what industry are you in? Or should I say what do the users mainly do?
Also, what kind of controls do you have on internet usage (matters for malware)? Mainly what kind of issues do you have to fix? Are the users spread across the country or all in a central location?
Basically the complexity of your architecture makes a huge difference.
tech. i have 300. when i need help we establish a project and bring in contractors
If you're running a distributed system where each node is exactly the same and you just push out a standard image then you could have a 1:1000 ratio.
But if you have a a bunch of computers doing very specific things each one being different the ratio has to be less.
An average doesn't really make sense unless you can specify the usecase for these systems.
...and that is all I have to say about that.
http://jessta.id.au
"Are we seriously understaffed, or is this normal?" It seems to me if you are able to read/post on Slashdot AND maintain those systems, you're doing something the rest of the world needs to look at.
"I can't give you a brain, so I'll give you a diploma" - The Great Oz (blatently stolen sig)
Was the ratio i had when i was managing an entire assembly plant's IT operations. Yes i was working my ass off.. 24/7 operation as well.
Where i am now, its more like 500:30000 ( ok, not a true ratio, but i wanted to include the total number too, since its pretty high. )
A lot depends on what industry you are supporting, your user base, and your budget.
---- Booth was a patriot ----
at my office I support 200 users/pcs, 10 servers, 4 copier/print systems, document scanning/indexing systems and warehouse production systems - and I'm not allowed to use GPO. my company has other locations with less users/systems to support but with more support staff...
the history of the world
I've modeled several models of this type, however there are broad variations depending on your environment.
Gather metrics on your work, build a taxonomy, build projections, tune the projections to fit new real world data, until you have a reasonable validated model.
Just to get you started: break your users down into light- medium- and heavy- support groups, and estimate how many calls per month you require, and how long these calls are for each type.
Same with systems, you probably have some hotspots such as specialized servers, or particular app/user combinations.
Break down the type of support tier required to service each type of request, three tiers is probably fine.
Don't forget 'devices', such as printers, that require regular and ad-hoc maintenance.
Budget increasing time for "training" users to perform standard tasks for you (such as cartridge replacement). This is ongoing, as you have staff changes.
Build a train the trainer model if possible.
Then shift this training into improvements in service intervals / rates for those areas.
You need to build departmental "emergency response teams" to handle crisis situations.
There are probably a few hiccups, like upgrades to Win7 if you have to, etc., these can be off-sheet, but need a separate burn-in curve for supporting user transitions.
Just an eyeball metric: sounds like you're running a call center or financial operation at 2.5 or 3 shifts (900 machines x 3 shifts = 2,600 users, plus the "low"). These users typically need about one person per 80-150 active users, however you need reserves (contractors, trained users, whatever) to handle large scale failures, such as patch Tuesday problems.
If I were staffing this "smartly", I would have 8 core staff across 2.5 shifts, have trained at least 3 people per shift to assist in functional stuff like printer paper/cartridges, and have an additional 4-5 contractors available "on call" for 1 hr response 24 x 7 as needed.
I'd like to encourage ongoing training : you'll find people who enjoy it, master the field, and end up happier, and helpful within your organization and in their lives.
You have 900 computers (not even counting the servers) and 2600 users? And you got 3 people "taking care" of all that? I assume you're not counting on these "low knowledge" users to help and/or support themselves or each other...
If you roll everything out properly and don't have to deal with constant dumb shit from people (e.g. the "Spyware of the Week" calls), then yeah I guess you'd be OK with that headcount. As long as these are not "power users" and you're located within an acceptable distance of all these machines. Like dealing with "System won't boot, unable to get into Safe Mode" cases over a long distance call?
We have ~2500 Windows machines (some with max 2 logins/users per machine, though most are dedicated to one person) and 7 top-level techs situated in strategic locations throughout the country. As well as a "Help Desk" team of 7 people supporting initial support calls. So that's anywhere from 200-350 users for each Help Desk tech and when necessary the higher-level Desktop guys.
Then we have a slew of other IT teams (SysAdmin/Network/DBA/etc) which I'm not even counting. Total IT headcount is about 25.
I can't fathom doing what you do with 3 people. Unless you're not telling the whole story.
Also shouldn't it be "Computers/Users per Admin/Tech" and not the other way around?
I hate it when I have 200 techs supporting one user. The cubicle gets crowded and fist-fights are not unheard of.
The minimum number of help desk technicians required is always directly proportional to the total number of ID10-T users they are to support.
"Be prepared, son. That's my motto. Be prepared." --Joe Hallenbeck
Make user an admin on their own machine if you want to increase the IT staff. Lock down group policies if you are overworked.
Trying to determine whether you are understaffed by looking at ratios of IT staff to users (or computers) is not the correct way to go. It is the simple (minded) way, but it is not the correct way. Picking a particular ratio makes many gross assumptions about your environment.
A better method is to review your workload: Are you run off your feet? Do you have to put in lots of OT? Are you spending your days simply fighting fires? If you found yourself answering "yes" to one or more of those questions, you are probably understaffed (I'll give you the benefit of the doubt and assume that you're not grossly incompetent 8^).
That said, three guys for a fairly large shop does raise issues like vacations and sick leave. Three guys might be enough under normal circumstances. But will two guys be enough? Or one?
linquendum tondere
We have about 400+ workstations with various OSs (Win 2000, XP, Vista, 7) and 20+ servers with various server OSs (Win Server 2000, 2003, 2008) and we only have the System Admin and a "sneaker tech". Oh and we have a remote location as well which they must deal with on the other coast of the US. At least from my viewpoint, your situation is not unusual.
At a hosting company I used to work at, we were severely understaffed. We had 4 techs to cover 24/7/365 and about 4000 managed servers, colo, dedicated, etc. It was hellish, but we got the job done.
we manage a cluster of around 5000 machines under Linux/Unix.
we are 5 person.
you do the math :)
6,100 employees 44 locations + 2 datacenters.
3 admins (1 network 2 systems) and 2 helpdesk technicians
We currently manage ~80 Linux, ~170 Solaris, ~150 AIX systems with a manpower of 6 architects/L3 (count me in), 3 admins/L2. Our SAN (half pB+tsm+CDL) team has 4 guys and for wintel/VMWare (+500 systems) park there are 3 admins L3 and 4 admins L2.
The larger the corporation the more per\user per\server to admin. Theseare my observation sover the last 12 years in my career:
For small corporations (less then 1 million) I usually see about 1/800 ratio for support\end user and 1/50 for servers.
For medium corporations (greater then 1 million but less then 80 million in revenue) I usually seea bour 1/2000 ratio and 1/150 for servers.
For large corporations (greater then 80 million) I see about 1/3800 and 1/250.
Support metrics are usually driven by "Call Times" including resolve times and hold times so depending on the scale of the businesses and nature it isn't so much support/staff ratio but rather hold time\support ratio. ITIL was crafted specifically to facilitate outsourcing Incident Management (password resets and all that less then 15 minutes crap) to lower cost, drill down labor and maintaining low hold times versus Problem Management which is the higher skill set.
Server ratio is largely due to "bucketing" of servers\apps to an admin resource (Think along the lines of an Account Rep). A.k.a Bob handles Apps A,B, and C along with Servers X, Y, and Z. So depending on the corporation you can have anywhere from 2-8 apps assigned to a single admin. Each application may maintain upwards of 5-12 servers depending on the size of the application. Smaller enterprises tend to have smaller "buckets". A typical LAMP stack may have 1-4 app servers, 1 NAS, 1 batch server, and possibly it's own database server. As you get larger those buckets share other buckets so you may have a team that handles just apache and another that handles just MYSQL\POSTGRES\etc. Those buckets can be huge. I have a team of 8 DBAs managing right now 2307 database instances. That is roughly 289 server instances per DBA. A simple table update may take 12 minutes for a structure update to process so median process time may factor into staffing requirements when concurrency isn't an option based on outage windows. Databases are virtual servers usually with a SAN hosted on hardware that is managed by another team but you can get the picture. By specializing administrative roles you can increase the nubmer of server or services supported by a person (power of scale) so the ratio of servers per tech tends to rise the bigger the corporation. In addition more expensive, comprehensive tools, become accessable to larger corporations (TIVOLI framework for instance.)
Based on your description you should need:
2 Call Center Incident Management crew
2 Problem Management crew
1 Senior Network Adminsitrator\Network Architect
3 Junior Network Administrators
1 of which is responsible for security\auditing
1 of which is responsible for maintenance
1 of which is special projects
All three should rotate these roles quarterly or annually as well as rotate 1 as a Problem Management staffer (the non-special project members)
So your total support crew should be about 8 people. You may also for off hours support want to outsource to a location 12 or 6 hours offset based on your location. (6 hours makes meetings more practical as you can usually get a meeting when one group is just getting in and the other is just getting ready to leave.)
-=[ Who Is John Galt? ]=-
I helped deploy a configuration of Sun's older Linux based Java Desktop System which allowed install, upgrade and configuration of 8000 desktops across hundreds of branch offices. It wasn't widely reported because Sun dropped this Linux based desktop to allow more focus on its own kernel and Sun Ray technology which easily allows more users per administrator. For Sun Ray desktops on Solaris I would imagine the ratio can be at least 10,000 to one.
It really depends on who your users are, what your servers are doing, and what level of support is desired/needed by the organization.
If you have 1000 users who use one app that they are all familiar with, user calls are going to be lower than an org with 20 different apps and a wide range of skills. In a large organization, you may be supporting everyone one from accountants who know Excel inside and an out to janitorial staff who are still using dial-up at home(seriously, they still are because they can't afford broadband) and have trouble using email.
Same with server apps. 100 file servers with the same os are going to take fewer admins than 75 file servers running a variety of operating systems and doing file, print, database, web, email and proxy/vpn/remote access.
Some orgs want techs to really know how the organization works so they can not only answer questions, but understand the business processes and be able to come up with new processes and services. They may also want that personal touch where you walk to the user's desk and help out. Other orgs are happy having basic phone support and making the users responsible for finding out the answers on their own.
While I was the sys admin for a small ad agency of 50 to 75 users (fluctuated monthly) my boss claimed that he read or spoke with other companies who were operating with a 50:1 ratio. Granted, ad agencies can be a bit unique with a very mixed technology environment and REALLY difficult users, but it worked for me. Most of the time I was fixing failed systems but I still had time to implement newer / better technology to resolve reoccurring issues.
I think what a larger company needs is a support staff that is busy 80% of the day with some good down time, and then a fair number of individuals who are working on more permanent solutions to the bigger problems. There are too many variables involved to give a definitive ratio of users:admins, but the work load is key. You don't want those guys staying late every day, and you don't want them sitting around bored either.
Probably the most effective thing to do would be to have them document their time somehow. There are apps to help with this. Guard against scaring them into giving you false information - sometimes employees fear that they are being watched and may falsify their claimed work load and you end up hiring people to compensate the inflated demand.
No sig for you. YOU GET NO SIG!
OK, not exactly, but the same topic was recently discussed in the Uniforms post.
(T>t && O(n)--) == sqrt(666)
Is the article writer suggesting there should be more admins than users?
As long as you manage expectations and, make management aware of the costs vs. risk.
There are a lot of factors that determine the manpower needed:
1) Are backups of files required by all users or just a few users (ie CEO, finance, hr etc)
3) What is the expected work hours 24/7? or 8/5?
4) How much redundancy is built into the servers? If they are in pairs then one can go down and be fixed during regular work hours.
5) How is your storage set up ? Disk drives fail so you will need RAID, mirroring etc
6) Have the servers been properly maintained ? A lot of times applications are put on servers without correct startup and shutdown scripts. If there is a power outage and the servers are cycled they should come up correctly with all applications.
7) Do you have a monitoring system in place to try to catch problems before they become catastrophic failures.
8) Do you have good configuration management so that the systems are as identical as possible.
9) Do you have routine maintenance items automated - this is easier if you have item 8 above.
All of these have cost and risk associated with them so it is important to go over them with management. You can explain how if you don't have a decent backup plan then data can get "lost."
- Things are the way they are because they're coded that way -
I used to handle about 400 hundred people and 400 boxes as a sneaker admin. So two folks for PCs and one for servers is workable -IF- you have your act together.
My suspicion is that you all are a bit low. Reasons:
- People get sick.
- People take vacations.
- You need to cross-train each other.
- Special projects can and do come up.
- There are under-met needs in the company.
To get the person that you need, you really need to show the business case for it. Once you can do that, ratios don't matter.
I've worked in a school for 2 years and they had 1 admin for eight thousand people among workers and students who used the school's network.
You are understaffed if you don't have the resources to complete the required workload. Otherwise, you staffing level is sufficient. If you are even asking this question, you are probably adequately resourced, since you'd clearly have noticed if there was a problem, and wouldn't be asking the question.
Trying to apply some general metric to determine how many staff you require is stupid. This would depend on a lot of different things: what platforms you are supporting (Windows clearly has a higher admin workload that most UNIX), what end user software you support, any infrastructure projects that need to be completed, what gets contracted out vs what's done in-house, ...
When I did desktop support we had 100-150 users per admin.
The load depended on the department.. The guy supporting the executive offices was at 50:1 and hammered most of the time.
You can push those numbers higher but then all you do is fight fires, projects don't get done and vacation coverage is a nightmare.
We came up with a chart at one point classifying each department as low medium, high and VIP(CIO,CEO,CTO, VP and so on).. Low was counted as 0.5 users
medium 1 , high 2, VIP level counts for 3. The admin supporting the nursing staff had 250+ real users but they don't put in many calls so it all balanced out.
For Server support I like to have 3 admins just so we have coverage for sick time, vacation and on call rotation.
It's hard to justify 3 with only 25 boxes so one could be a dual role as desktop lead/server admin.
It's always nice to have a guy playing desktop and server to keep communication open between the groups.
If you think it's expensive to hire a professional to do the job, wait until you hire an amateur. --Red Adair
This was actually a recent topic over where I'm working at and we were trying to figure out a way to explain that we needed another person. We went the route of trying to find this magical industry average and like everyone else is saying "it depends". For our level of support and configuration here - 1 admin per 75 windows workstations seems to be the magic number. Right now we're at about 1 per 150.
However, "industry average" doesn't mean jack to the client. It's kind of like saying to them "well Joe has a Ferrari, therefore we need one too". The spenders want measurements (aka: "metrics"). For IT guys who are used to instantaneously jumping up to fix every problem the minute it comes to their attention - this will drive them nuts: what you'd need to do is measure your response time and time certain tasks take. Show them how many resources you manage. Our client couldn't believe we were supporting 300 machines. "I thought we only had like a hundred". They also had no idea how long it took to set up a new computer for a user. They figured it to be relatively instant, no factoring the time to carry it over, to tear down old equipment or clean off a table, add the user accounts, run the location's security patches, transfer old data, to build the initial image, etc. We found out we take an average of 3.45 hours per machine for an upgrade (moving gigs of data and making the master image averaged out per machine were the main hogs of time).
So "it depends" really applies to your question too when asking "what are you trying to say with this information?" Are you trying to justify the addition of a new employee or explain why response time has dropped significantly, for example? You'll want to be able to paint a picture that'll relate to their dollars. You'll want to be able to answer this when they ask "Will my decision to [add employee, upgrade software, whatever] affect cost (dollars and man-hours), scheduling (time to complete/react to situations), or performance (quality of product/service)?"
Rather than looking at ratios, shouldn't you be looking at work load completed per tech per day.
The ratios are going to change depending on culture, type of business and products supported. The more independent your culture is, the less techs you need. The most products supported, the more techs you need. For example, we support over 500 different applications, that makes it much more difficult to support than just office and windows; again that changes your ratios.
Look at how many incidents are reported per day, how many hours are spending for root cause analysis, how long does the average server outage last, how many users it impacts, etc. Add that all up.
If your not in a spot where you can measure this type of time spent, rather than worring about being understaffed, you need to worry about the documentation of issues in your environment first.
All other estimates are useless.
One place I worked in the 90s, about 100K seats, about 10 HD on the phones, about 15 field techs, about 5 WAN guys (myself and 4 others), about about 10 operators / system programmers / unix guys, works out to about 1:2500.
This was an IBM mainframe shop handling about 5% of all retail stock transactions worldwide on "dumb" terminals. If they didn't demand 24x7 coverage to handle worldwide markets, it would have been somewhat cheaper as we often staffed for "just in case". Having someone on pager was unacceptable per the SLAs and marketing, so we had to be on site, so they had to hire more people, even if all we did was read technical manuals all night (IBM had some pretty good textbooks for ATM networking, assembly language, COBOL, and a few others)... The IBM onsite engineer mostly did hardware work, everything was triple to quad redundant, and that redundancy was often tested and the customers never knew...
I guess those pesky PCs need one per 100 to 300 PCs to get an "acceptable level of service", for a PC anyway.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
So I've seen 1 sysadmin per 30 Windows Servers and 1 tech per 900 workstations. What do you do on patch Tuesday? How do you roll out these systems? Are the workstations solely running MSOffice and IE for a single homegrown webapp and nothing else? Does your "30 servers" include 1 SBS DC and 29 file servers with 0 printers?
I think before taking wild guessing and "epeen" statements of "I has less techs than you and we're flawless!" you're going to need to figure out what quality of work you want.
Are all of your users running as admin? No problem, I'm sure you've educated every single one of them about updating their software, avoiding Adobe exploits, staying off of social media sites and license compliance. I'm sure they do all of their own asset tracking, too. But that's clearly not going to work at an educational institution worth any salt.
How stable is your environment? Can you deploy a single system with firewall and reasonable security that has all the software a user will need for the next 3 years? Or are you in a "normal" environment where they need updates, new software and new hardware swapped on a regular basis?
1 tech per 30 servers, 1500 workstations to serve 4,000 internal customers 24x7x365 is ridiculous and absolutely not possible in a business that has to deal with compliance (PCI, SoX, HIPAA, FERPA, etc). Unless, of course, it doesn't matter if you actually stay in business and/or out of prison....
I've heard that Google runs 800 servers per SA - corrected to read *one* server and 799 clones, per SA.
I don't envy you, sir.
That's kind of funny... 900 CPU's (clients) and 2600 users for you? At our place of business, we have much more like a 2:1 ratio of CPU's to people... Not counting VPN connected foreign hosts... How does it work out that you have fewer than 1:1? Are you leaving out dumb terminals and network connected embedded systems?
The answer would vary depending on the complexity of the network. A semiconductor company with mixed win/lin network and lots of different types of servers will likely to require many more IT personnel than an accounting firm of the same size but more "homogeneous" network. That's probably why there isn't much statistics on the average tech per user/computer.
When I was there, RUCS (Rutgers University Computing Services) ran a campus of around 18k people from four active student admins and eleven never-seen never-active bureacratic staff. So that gives a ratio of about 4500:1. It's worse if you consider time of day, though: that was three during the day (6000:1) and one at night (18000:1).
Disclaimer: it's been almost 15 years, the school's size may have changed, and there's an outside chance that RUCS might have gotten rid of their dead weight and gotten their shit together.
StoneCypher is Full of BS
I work for a small company with a relatively big IT system. We have about 250 people; 6 servers (Win2003/2008) and ~300 workstations (XP/Vista/7) over 4 locations. We have 2 full time staff - the IT director and myself. We are only slightly understaffed: we fix petty crap and teach classes to the point where we would have to work some overtime to take care of everything. Solution: we outsourced the "serious" network/VOIP tasks to a local firm (also a small biz) so that we can focus on building relationships with the rest of our company. I've been with the company just over a year and it's working out really well, though I would still love to have a P/T person to work tech support. I'm not sure why, but I love teaching.
As others have said, it really does depend on the work-load.
Some have suggested at 1:75 is "low" - for me, based on the workload my staff have, that's *high!!*
I have 4 groups supporting various workloads in various geographies. /number/ of users we support isn't the issue - it's the workload those users generate. Some part of that is due to the types of applications we support at each site. Some of it are just the demands some of those departments put on us (e.g. continual last minute requests). Some groups are highly self-sufficient. Others rely on IT day-to-day to help develop solutions to increase their own productivity.
They range from as low as 1:20 to as high as 1:70
For us, the
I've asked the same question the original poster asked for years - what's the "right ratio?" The answer is: it's unique.
Hello, I cannot give my numbers but I'm pretty damn sure my site is incredibly understaffed, so I'm really curious on how it is done on other sites. I know this article is touching the desktop side of things, but if you can give some input on the ratio for admin/servers, it would be appreciated.
We have 1 central office, and 61 schools across the country. All technical support comes from our central office. For the client aspect we have 2 system administrators, and 4 Helpdesk technicians. We have 2,400 Windows 7 desktops/laptops, about 7,000 thin clients, and 72 Windows 2003 terminal servers. We are forced to design our systems and infrastructure to be very hands off in order to support these numbers. We automate everything we can.
You need 1 IT staff (helpdesk, sysadmin, etc.) per:
20 Windows servers
50 Linux servers
100 full time computer users
1000 part time computer users
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
Yes, you are seriously understaffed. And yes, it is normal ;-)
My wife doesn't listen to me either...
I would say you are seriously understaffed, and that is completely normal. I.T. sucks. I got out of that field to pursue electrical engineering.
is, or was, the industry standard when I was still managing a help desk a few years ago. This assumes a typical windows environment with a half-way decent standard load and other people doing your
infrastructure and server support.
We have approx 6000 users and about 5500 computers. 70ish servers (not sure as I'm a network admin) and 213 pieces of network gear (routers & switches mostly, maybe a dozen firewalls).
All this spread out around 35 remote sites.
We have a core team of 5 admins - 3 server geeks and 2 network geeks.
There are 4 desktop geeks who do the remote sites and user support.
And finally 3 helpdesk staff who field calls all day.
And most of us have very little on our plates these days.
Well I work for a school board, the team I work with is 5 techs + 1 Director, we all cover pretty much all areas Desktops, Servers, VOIP, Video Conferencing. We have approximately 2500 workstations, 80 Servers (mostly windows), 1000 full time empoyee and 6500 Student users. Are we understaffed, Hell ya, do we get the job done, damn right!
I See a lot of commenters stating that it depends, but not depending on what. In our case we can get away with our staffing levels due to the technologies that we use to secure our machines from users, Deep Freeze, with out it our staffing levels would need to be much higher because we would have to deal with virus/malware problems on top of general user mucking about. We use SMS to keep our machines up to date in the afterhours. We also like to label ourselves as control freaks and that also helps.
It really depends on where you are to what you need. Do you have people who never want any other software installed? Or do you have people who -need- various programs or want some to get the job done easier. Do you provide -all- your own hardware, or do people occasionally bring in faster printers so they don't spend all day printing? What about viruses? Are they a big deal? If say, 25 computers go down in a day due to something like a power surge will that seriously affect productivity? Are these systems running similar hardware or are they a mismatch of various cheap systems? Will these systems need hardware upgrades? If you have systems that are really static and won't need many more things done to them, yeah, 1 sysadmin and 2 sneaker techs might work. If you have systems that are dynamic with software changes, hardware changes, or computer systems that aren't identical, it could really help to increase the number of sneaker techs. In most cases one sysadmin is all you really need, but bumping up the number of decent sneaker techs you have really helps increase productivity, few things are more annoying than trying to fix someone else's machine and then someone "higher up" thinks you should drop everything and fix their computer (when really in reality they don't need their computer to begin with....).
Taxation is legalized theft, no more, no less.
How many admins does it take to change the light bulb in the monitor? I would say more depends on your environment as far as the applications, level of demand, tolerance for downtime/repair time and responsiveness than on the simple number of systems you support. I'm the IT manager for a company with around 500 employees, but only about 150-175 of them are really computer users as part of their job. We have 30+ servers in the "admin" network and another 12-15 in the departmental network that supports out prepress operation. We run monitoring, process control, list processing and logistics systems, as well as the range of business applications and financials common to any organization. My current staff is two, including myself. The main thing is how well your processes are constructed, how well monitored for pre-emptive response, and how well you manage the time and resources to fill the support needs of the company. No formula of X number of techs for Y number of users, Z number of computers and K of servers is going to give you anything like an accurate estimate, because it lacks the essential parameters for how much work the functions you are required to handle actually require. Someone else stated it fairly succinctly - if you are able to meet the needs, you have enough people. If you don't have time to support and monitor the systems because you're too busy putting out fires, you could have 1. too few people, or mosre likely, 2. poor processes for getting the job done and staying aware of the conditions that cause the problems in the first place. Its often easier to do the same thing more than do the better thing. It takes more insight into how to get the results the best and most efficient way to improve the process and get it done with better equipped, trained and directed staff. I'd start with a review of what gets the job done better, and staff for that.
My rule-of-thumb is that there should never be fewer than 3. The reason? Vacation, sickness and on-call.
With one person, vacation gets difficult. With two people, on-call gets difficult when one guy is on vacation - in addition to the problem that a person might get ill while another is on vacation.
How many machines can 3 people handle? That, again, depends. If it's a standardized setup for all servers/workstations - with all techies knowing how to handle all of the systems - then you probably do not need too many people handling it all. This, of course, also depends on the expected rate of change.
On the other hand, if you have loads of different systems, either on the desktop or server-side, then you'll need more people handling it all - since it's rather unlikely that everybody will know everything.
A couple of examples:
- For a large cluster of machines, say 5000 - all running as one cluster / supercomputer, you don't need very many people (unless there is a huge rate of changes). Why? Because it's all redundant and you've of course developed automated installation routines etc. for nodes.
- With 500 different servers, with different OSes, different services and hey - different customers, you'll need way more employees. Quite simply because of many different people contacting you, bigger rate of change, more systems to know, and so forth.
At a Fortune 500 company where I worked a few years back, the user-to-IT ratio for Windows users was about 4 to 1. On the Macintosh side it was about 100 to 1 although a 50 to 1 ratio is probably better.
It's the same as any other department in your organization. If you have one sales guy and he's working 60 hour weeks every week to maintain the status quo, then maybe you should look at hiring another sales guy. Same thing for IT. A lot of it will depend on how much can be centrally managed/fixed from your desk/PDA and how much time you actually have to go and spend helping users one on one. We have about 90 users and 25 servers with one network guy (me) that also handles user support, one dba/app developer, and one manager that interfaces with the rest of the org to allocate our resources. If we were able to hire someone else, it would probably be another developer because that's where our shop sees the biggest backlog of work. Our developer spends most of his time maintaining existing apps, so he has very little time to develop new apps. I think the biggest factors that affect the answer to your question are 1) user skill levels 2) your power user's willingness to help their coworkers 3) having a centralized helpdesk/issue tracking software where users report problems and you keep track of everything that needs fixing and prioritize, and 4) your ability to fix problems from where ever you happen to be (office, home, Blackberry, iPhone, etc). One thing I've found helpful is interns...many times you can call down to your local community college or university and get setup with their work study program. Most of the time the interns they send you will work for free, and they're perfectly able to Ghost systems and remove spyware, freeing you up to do more important tasks.
Well, the number I usually hear in the industry is 1 SysAdmin per:
10 Windows Servers
100 Unix/Linux servers
I think that the numbers are little low, and maybe too idealistic...Depending on the application, I could *easily* see one competent SysAdmin managing at least three times those numbers. Of Course, that is the rub, a competent SysAdmin. A *good* SysAdmin could probably increase the original numbers by a factor of ten. But then you start running into the "run over by a bus" syndrome...if a *good* SysAdmin can do that, what happens if they get run over by a bus? Many times companies take too much advantage of a *good* SysAdmin, and then when they leave/burnout/get hit by a bus, they spend months (years?) trying to replace that person with just one body, when the reality is that they may need two, three or more people to replace a *good* SysAdmin. I watched it happen at a company I worked for...when I left, they kept on burning out SysAdmins, and every two or three months, they would be advertising for a replacement SysAdmin. Oh, yes, I was also doing development work in addition to my SysAdmin duties. A Co-work's opinion was that it would take at least 3 people to replace me.
ttyl
Farrell
CAN-CON 2019 - Ottawa's only book oriented Science Fiction Convention! October 18-20, Sheraton Hotel, Ottawa, Canada h
I admin a webserver that currently averages about 2,500 visitors at any given moment. Since pageviews pay the bills, they are my users. Their problems quickly become my problems when their problem is blank pages or "oh snap, the database is down!" errors.
Ugh.
The only advice I could possibly give to anyone with this sort of situation is to invest in a respectable high availability setup. You know you're doing at least something right if you wake up to find half your farm down but analytics is still showing business as usual. My setup is currently lacking redundant db backends and backup email. I'm very seriously considering moving email to gmail but thus far the only issues I've ever had with email has been my own stupidity.
On the upshot, my troubleshooting consists of a copypasta email along the lines of "clear your cookies and restart your browser", which solves 99% of the problems my 'users' have. If it doesn't, then it's an obscure problem on my end. Most people that have had issues, if they care enough about the site to report it are usually quite willing to help me debug the issue by providing screenshots and such.
Education:
5000 staff, 33000 students
8 sneaker-net type techs at a central location, 1 tech per high school (4 HS), most middle schools (7) have their own techs as well.
3 sysadmins (one of those is me)
3!!! helpdesk
3 - 4 network techs, depending on if you count the manager who does quite a bit hands on
8 application administrators for things like oracle or our student info system
NO DBA's, all on contract, and we run Oracle for finance and have about 15 sql instances, which is undergoing consolidation b/c the folks running it before let it sprawl...
13,000 workstations, 120 servers (netware/windows/linux - 45-45-10%)
Am I understaffed? Yeah. For a time we just had 2 sysadmins and let me tell you how much fun that was.
Another site I worked at had 4 people at HD, 2 server admins, 1 network admin and 2 or 3 sneakers on each of two locations. Roughly 600 users
I'd say it's all about the number of service requests you get and the complexity of the IT environment and not absolute numbers.
Add SLA's and what users/management expect to determine if you are understaffed or not.
Oh, and another thing...most employees operate with impunity when it comes to getting their PCs all spyware'd out. If you can change that, it will help drastically. If you find that you're wasting a lot of time removing spyware and cleaning up the mess from webshots and all the other crapware employees like to download, I think you'll find that your spyware volume will fall off sharply if you can get management approval to start charging your spyware cleanup time back to the employee's department. Once the department head has a talk with their employee about how they don't want to see another 5 hour charge for removing Troj.Vundo, the employee will be much more careful about what he/she chooses to download at work.
Well ..if this is helpfull for your statistics , im my work , we are 3 sys admins with around 100 linux boxes and 30 000 - 40 000 users.Their are 3-4 support ppl helping out.
So i guess out ratio admin/lnx is 1/30 and admin/user is 1/10 000 - 1/15 000.
The number of admins needed in any shop depends on many factors, especially automation and duplication. In an ad-hoc environment where users are given free rein, you will need lots of admins. If management will support restrictions on users, the admins are creative and the necessary tools are purchased, then the job can be handled by a few astute individuals. If anyone knows of such a shop...
-- Consensus - 50% probability that the majority are wrong.
at least 1000 servers per admin.
I've worked for my current company for almost 8 years. I am the only IT staffer and currently support 75 desktop users (65 pcs, 10 Macs), 8 servers plus 15 more virtual servers (all Windows), including the a web development sandbox, the PBX, the security system, and all the network hardware (mostly Cisco).
I used to have an assistant/junior IT person, but I had to let him go last year when the whole company lost 40% of its staff.
The only help I have is from the web developer, who has some knowledge, but who has gigantic holes in some areas- especially networking.
Are you busy? Just because you've posted the question doesn't necessarily mean that you are. Bodies usually are warranted on a need basis not based on how many of X that you have. In a small shop you're probably going to have a mixture, like I do. We have 110 Employee's, 12 Windows servers(2 DC's, 1 SQL, 1 X-CHNG), 6 linux servers(2 are XenServer Hosts, 1 Firewall). We both handle everything, neither of us are assigned to just do Servers and the other to answer the phone. Honestly we hardly get phone calls, and when we do the Tech Support dice could answer the question(no joke). Usually simple stuff. I've been more busy with server maintenance than anything.
Anyways, the point is how busy are you? Is some of the slack not being picked up by other tech's? If I was also judge this based on machine count it would be around 50 to 60 machines per Tech. Probably around 500 Machines warrants the segregation of Tech's(phone and email support) and Sys/Network Admin.
Sometimes, the answer is to just destroy it all.
Unfortunately I don't have the specific links, but there have been some studies of IT budget relative to revenue with different baselines for various industries. Using these, I was able to argue the point with the Management team that our IT department was grossly under budget/under staffed, even when compared to the lowest baseline for all industries on the chart (each industry had a different percentage range - low for manufacturing, high for "tech" industries").
It is not exact, but is the closest thing I could find that gives any quantitative data related to this subject. I specifically looked up "IT budget vs. revenue".
As a former Regional ISP with full email, webhosting and News services, we started at 1 tech for 150 users in 1995 when we had to talk each user through the install of trumpet winsock and ftp download of Netscape (or physically go to their house and do it). Later this moved to 250:1 and this included supporting Windows 3.11/winsock and Win95 on 4MB of RAM plus issues such as how to click a mouse, what is a modem, my printer won't print, and Packard-Bell (need I say more). With WinXP and DSL, this moved to about 500:1 and still includes support on reloading Windows, printers, virus detection and removal. Not to mention, much more involved mail and web hosting issues. As for admins - Telecom needed 1 per 2000 dialup lines or DSL users or 300 hi-cap circuits, Network 1 per 100 dedicated customers and SysAdmins at 1 per 60 servers if they run a broad range of services or 150 servers if they run similar services or 500+ servers if they are identical services. Windows versus Linux doesn't make as much difference as it used to because Windows has fairly robust scripting available if you learn it.... but in general, Windows will up your admin needs 30% because few people know how to use extensive scripting on a win server. Current ISP's generally limit support to services sold and need half of less of the resources. The bottom line, resources depend heavily on exactly what you support, the tech level of the user, and how diverse their hardware is.
I run two isolated networks for roughly 300 users (about 600 accounts between the networks) over multiple isolated geographical locations (you have to convoy between them) with 2 admins and 5 sneaker techs. We run about 200 computers, 4-6 servers, and multiple wireless interconnects with associated equipment. Of course being military makes for special requirements.
I'll meet you at the intersection of "Should be" and "Reality"
1. Does the administrator have complete control?
2. Can the servers/desktops run effectively as diskless or thin clients?
3. Are all servers installed from the same base image/jumpstart/kickstart?
4. Are patches tested and maintained across all platforms on a regular interval?
In a perfect world, all 4 are true. Under those conditions it is possible to manage 200 UNIX/Linux server systems per admin. Given thin or diskless clients, the number of desktops supportable per admin is very high.
The worst case scenario is when all systems are different and carelessly maintained. The number can drop to 25 per admin on busy/active systems.
We've got ~2800 RHEL4 and 5 servers and ~400 VMware ESX servers and 4 admins. The key is homogenity.
Gentlemen, Ladies, Techies and others------
I have had the honor and problems of being an admin for a relatively small user base of around 90 users and 3 servers as well as a 16 node Linux cluster for advanced computing and analysis. I have also been an engineer and engineering manager in a company with 60,000 users and have had to deal with many of the issues discussed here, particularly in last 2 years with the recession and RIF's. Here are some comments, helpful or not, that I would like you to consider, now that I have read many of the posted comments.
1) The IT department, in all but a few cases, is a SERVICE for the rest of the 'productive' employees. Meaning that you don't work on or produce the end items that make the company profitable. In that capacity, your duty is to provide what the CUSTOMER (user) needs to perform their job in the most effiecient and timely way possible. I have often read about 'Locking Down' users computers. HA! This is particularly annoying and should be avoided. YES the top OS areas should be protected but preventing users from accessing the clock is a bit much and when traveling with laptops becomes quite an issue when Outlook thinks is 3PM when it's really 1AM where you're at. Also, in many businesses you need to experiment with trial software to see if something is possible before ordering 100's of copies. In engineering fields this happens regularly. Just downlaoding a new template for Word, Excel, or Powerpoint becomes an issue when locked down. Inability to use flash drives, inability to update graphics drivers etc. become a Herculean task and delay the completion of projects. IT regularly pushes updates that SEEM benign but suddenly you have a flashing display every time you open Word or the fonts don't look right. What is the basic fix at most IT dept. Send in a ticket. The ticket goes round and round and where it stops no one knows. At least 24 hrs before you get a casll on something you can fix yourself with little chance of mistakes.
2) Most of you IT folks think about it backwards. You should be 'out there' with the users. There's nothing better for a smoothly operating network and business that uses technology for succes than well trained USERS. How do they learn the little things like how to check their own settings for bypassing the proxy than for you to sit down with them and show them how you diagnosed the problem and what the solution was. YES, it is EASIER to just do it yourself but this all stems from the fact that you have been trained as an 'IT weenie' that users are dumb and you should be able to do your job of serving hundreds with 1 guy.
3) You need more people!!! you need somehting more like 25:1, like a class room, however instead of playing solitare when you're not busy, spend your free time walking around cube city talking to your users. Find oout what daily issues they have that simply don't report (100's). Also, spend time making good desktop guides for all those issues that pop up and show everyone how to fix them step-by-step in a video or powerepoint (TRAIN the user).
4) You should constantly being checking your frequent updates on regular machines delpoyed normally. Every few weeks there is a pushed update of some patch or app update that screws somehting up and then we get 'service outage' emails saying they're fixing this. This didn't happen at my previous company, they meticulously checked everything, and it didn't take years, they added people to do it right and fast. At my current place, we are 3 to 4 revisions behind the times on many critical apps. We constantly have to strain to work with customers who are behind themselves but ahead of us. Then we request updates and 'IT' does selective update for those working that program which of course now means we can't communicate with anyone internally without significant issues. Who's thinking ahead? I know this is expensive but for a multi billion dollar company this should be in the profit plan and forced to be implemented.
I summarize by saying that IT is a service. The us
I think that the staffing level has far more to do with the systems you support than the sheer number of users. I have been in several shops, and the workload really depends on how much attention your applications require.
In the shop I manage, I have three geographical locations, 150 servers, 1000 users and only two sysadmins. Granted, we do have a dedicated network engineer and phone/telecom admin. But only two admins for all the servers, applications, users and workstation support. There is also a dedicated helpdesk team- but they don't really do any troubleshooting. We are a bit understaffed, and the ticket queue is a little deeper than I would like but we manage just fine. One more FTE would be very welcome, but the budget just won't allow it.
In my opinion, many shops are way overstaffed. If your staff has time to hang out on Drudge all day then there's simply not enough work. I hit slashdot once every couple of days, and my admins are engaged all day long. An earlier post said "6 server admins for 200 servers". That's nuts. Six servers out of 200 are broken every day that take all day to fix? These 6 people do nothing but fix servers? I'd layoff four of 'em, replace all the servers with updated hardware/software, still have a bunch of money left over and ask for a raise.
It depends on how many Windows desktops you are able to replace with Macintosh OS X, Solaris or Linux. Seriously. Windows isn't around because of it's technical merits.
I have worked in help desk environments in the past for a Windows / Macintosh / Solaris computing environment. The Solaris users largely took care of themselves, but contacted us for some settings information, like establishing the right settings for Kerberos, LDAP, AFS, or SMTP. The Mac users outnumbered everyone else by at least 4 to 1. However, it was the Windows users that wasted about 80% of our time for drop in help. Even cloned setups on identical hardware had different problems. Drivers were a big one. For phone calls, it was a bit higher in number of Windows user contacts but a bit shorter in duration for each one.
I did family tech support for years until I had enough and bought anyone who was willing new Apples. Only my mom took me up, but her support calls dropped off to nothing within days and now we can talk about other things for a change.
I've visited and toured libraries and schools using LTSP. One of those was stuck with some windows machines. The effort to keep the few Windows machines going was about, from their statistics, about 14:1 compared to LTSP. That ratio would probably been higher if they had even higher ration of Linux stations. The others cited even more favorable rates.
Getting rid of Windows is mostly a psychological problem. First, users have to become familiar enough with computers to be able to do their daily tasks. Having knowlegeable staff on the spot to nudge in the right direction is essential, as is encouraging peer support. Then they need to keep access to the Windows machines and try to do on Windows what they can do on computers. Then they eventually decide on their own, 'fuck it' regarding the Windows use and drop it without looking back.
The real question is are you always constantly working your ass off, fixing stupid problems - and therefore unable to do anything more productive? If so, then it seems you don't have enough people.
Setting the 'right' staffing levels, then depends on how much you can clean up the computing environment. I for one am offended that so much money and time is wasted just trying get the M$ stuff to work as well as its competitors. I would much rather see the same number of staff hours used not for support but for improvement and making things faster, easier, more productive. Before Windows, IT used to save effort rather than a live demo of the Red Queen's Race!
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.
A school I worked at: - 1500 users - ~1000 computers - 3 techs Other place I worked: - 100 users - 5 techs Now: - 70 users - 3 techs
The ratio depends on at least three factors: a) How organized you are b) How mission critical the systems are c) The degree of homogenity your envirenment allows you For example I have worked in a research environment with AIX desktops and servers, and 6 people were easily able to administer 2000+ systems due to the fact that we handed out local admin passwords for those who were willing and able. When I was working for a high profile company, we were 3 admins for 100 mission critical HP/UX boxes. Both environments allowed for a high degree of efficiency due to the homogenous systems. OTOH I've seen a 12 men sloppy IT department with mixed Windows XP and Linux desktops and servers, having trouble keeping up with 250 users at a local University...
I used to work at a telco and 2 of us supported about 100 users. It got WAY out of hand because they would just call us or email us if they had a problem. There was no way to prioritize the work that had to be done. If you got a call in the morning on a problem that would take most of the day, and then got calls for things that were minor (in terms of TTR), a lot of time was wasted. I now do basically the same thing at another company and we use a ticketing system. We rarely, if ever, run into those problems. Even if we did, we'd have more time to deal with it, because everyone MUST put in a ticket when they have an issue. If you aren't using a system like this now it will require that you talk to some higher-ups about getting a system implemented and there will be policy change at the company, but the people that you are pushing the idea to should immediately see the benefits.
Jeff Rothschild, the vice president of technology at Facebook, said in a recent presentation that Facebook has 230 engineers supporting data for more than 300 million users. He says Facebook seeks to maintain a ratio of one engineer for 1 million or more users. Facebook is vague about exactly how many servers it has, saying it's "more than 30,000." But at 30,000 servers and 230 engineers, that's a ratio of about 130 servers per admin.
Microsoft says it has automated its data center operations to the point where its admins can each manage between 1,000 and 2,000 servers. That matters, as the company may pack more than 300,000 servers into its new container data center in Chicago. It expects to support that facility with about 30 employees, including admins and facility maintenance staff.
RichM
Data Center Knowledge
We have one administrator for 80 users/computers and 15 servers at two locations. He does all the server admin stuff as well as desktop user support. There is a VP above him who can cover the user support when he's out of the office.
For example, in Soviet Russia, users administer you!
In my current shop, we manage close to 400 os images, about 200+ servers, > .5 Petabytes of Storage, > 20TB of backups/day and have only 5 FTEs and 2 contractors.
A lot of the FTE to Server/User ratio depends on how organized your shop is. If you have managed to automate most of your SA activities, it's not unusual for 3-4 FTEs to manage a 1000 server shop.
The key words are:
1) Standardization of platform (have at the most 2 or 3 platforms, chosen for specific roles)
2) Standardization of OS/Build (have 2 or 3 OSes you can manage well, eg: Solaris, Linux)
3) Standardized Architecture (define your architecture and stick to it -- SAN, TAN, etc)
4) Automation tools (such as a good monitoring solution, Fault Management system, etc)
5) A good analytics tool (such as splunk, which will simplify your life with automatic log indexing, parsing).
In an organization with centralized resources (trends have been gravitating towards centralized infrastructure, datacenter consolidation etc), this becomes easier. In a distributed IT shop it is harder.
What you described is almost exactly like my university - about 2800 total users, managing around 900-1000 computers. Lots of servers. This is what we have: 4 "systems support" (desktop, laptop, user stuff) 4 network/server/that kind of thing 3 or 4 help desk 2 or 3 for telecom about 6 for remaining tasks: web, software, management, etc. Now, we also have about 16 part time student workers for systems support, 4 for network/server, and 8 for help desk. I can attest that at times the 16 is overkill, but most of the time they are fixing small problems.
We basically have a 2 person IT department (including me the director) with a programmer that comes in part time 30/hours. This is for 200 desktops, 22 remote offices and about 30 servers (mostly linux). We also do a good amount of development on both web applications and in house resource management software. We wear many hats, and if your not taking a support call you are trying to bang out a program or design the network or find cost savings. As the director I put in plenty more than 40 hours and get my hands dirty all the time. Our user base is pretty low on the scale of computer skills which makes everything more difficult. We manage alright but we have a good back log of programs the company wants created, I just tell the CEO if you want us to go faster send money, if you would rather save money and wait we will continue as normal so that way he dictates it not us. We had 4 full time staff and interns when I first started but since 2004 we moved almost all our servers to Linux and have been trying to automate as much as possible thus our efficiency has increased and we have been able to pull it off. I think we are producing about as much as you can with our situations, I don't think we could get much smaller and still function as well. At one point in time we also managed all the email but we moved to Google apps about a year ago and that has help remove more pressure and thus I reduced the programmers hours.
If you manage it correctly and have some luck (as your idea's work out), you can get pretty small but it might burn some people out. I can say we are riding the fine line right now, but we are still considered the best corporate department and the field loves us and I use that to gage my decisions.
Our company is a cost cutting mode so that dictates a lot of my decisions.
I've worked in shops with 50:1 and 500:1. The quality of users, the quality of hardware, and the quality of the staff all factor in.
In the 50:1 shop, the machines were bottom barrel and with ~200 machines in the wild, they were failing at a rate of about two a week. Add another two failures because of the low quality of the users(did I mention they all had local administrator access?) and you can see where this is going. Also, nobody bothered to use a network based auto-patch system, so each machine had to be manually patched. Furthermore, most users had their own cheap printer, which added another fail-happy component to the mix. Don't even get me started on how bad it was when we moved from NT 3.51 to NT 4 - Manually installing NT 4 on 200 machines.. ugh.
In the 500:1 shop, all the users were pretty easy going, and none of them had full administrator access to anything. All the machines were auto patched via the network and support was usually provided via remote desktop. Oh, and the printers were department wide with support contracts.
Point is, there's no magic number to make it work. Streamline your efforts as best you can, and use a trouble ticket system. You'll find out really quickly how many people you actually need.
This depends a HELL of a lot. Are there tons of change requests? Are you constantly adding users, file systems, etc? 100:1 is OK if there are minimal changes, including patches.
In a test or R&D environment, sometime there are as few as 10:1. But the changes are constant and daily.
I guarantee that at 900:1 anyone can break into your infrastructure.
2 Managers 2 Techs (1 part time) ~15 servers ~250 users But we outsource/leverage quite a substantial amount from other business units in our platform. (Basically anything software wise is leveraged, we handle site-specific software and 95% of the hardware).
I'm a solo-admin and support center for:
20 in-house employees/PC's
7 servers (5 virtual)
160 field staff (only website support)
24/7 "on call" with only a "when we call you" stipend
The bestest part is that I report to two managers. I've got my full time manager (the office manager who wants me to leave her alone) and then the company owner when he feels like micromanaging someone. We all take turns "being incompetent" in his eyes, so every couple months he rides someone ruthlessly for 2 weeks, then another person for 2 weeks, etc.
The United States Army teaches that you need three to one odds to attempt an offensive maneuver....
For some CAD setups it can be up to 7% of the project's drafting budget.
At least 1774 servers with 8 Admins. So we're at about 222 servers per SysAdmin. I think that's higher than average.
It really depends on the industry, the shop, management, and users. However, when I was a server/desktop admin at an architectural firm, there was a book published every year that gathered lots and lots of IT management stats from various companies in the field. It encompassed staff sizes, experience levels, money spent on hardware, money spent on licenses, etc. It was extremely granular and gathered data from industry leaders as well as average and below-average firms. You could then do your own metrics and compare them with the published data. This was very useful when it came to dealing with management. I wish I knew the name of the book. There may be something out there for your particular industry, but you'll need to look around and talk to your counterparts at various IT shops within your industry. In fact, there was also a yearly conference where you could network with other IT personnel within the architecture industry from around the country. My manager found it extremely useful to hear how other shops were solving problems, and sometimes to simply hear that we weren't alone in our challenges.
... to me, the big questions are:
A) Is your backlog growing hopelessly longer?
B) Are you working too many hours every day (>10)?
C) Do you get requests that your team doesn't have the expertise for?
If the answer to any of the above is YES, then you probably need additional help. A manager needs to be able to recognize when things are under control and staff accordingly.
I wouldn't accept a job where I'd be the only admin. For one thing, the company won't be set to handle my vacation time or any sick leave in a way I'm comfortable with. More importantly, I find that being the only guy who works with something gets old fast. I want someone to discuss ideas with who actually understand what I'm saying and can provide valuable input.
The ratio isn't really what's at issue because that can vary due to the business you're in. What's the bigger concern is that the IT staff overlaps that's what I draw to be the biggest problem where I work. Our system support staff is responsible for end user support, network support, telecom support and server support. However: Telecom support is only responsible for telephones Network support is only responsible for networks Server support is only responsible for servers That's find and dandy for the other departments, but for us poor shlubs in system support we're overworked and underpaid. The CIO of our company said once that he expected at least 900 or so end users per IT staff. My location alone, I support around 200+ users with varying levels of support from co-workers and people on other teams. This also includes our servers, our phones, our networks.
I've worked in two companies - one had 500 users and 30 IT staff, one had 900 users and 10 IT staff... Number of servers pretty much linearly scaled with the number of users. Let's just say that in the second case, it was not enough.
"By many he probably meant the standard definition as used by the English language."
By many what? Heh.
Damping absorbs vibrations. Dampening is caused by moisture.
does it take to plug in a network?
"There are around 2600 users of varying knowledge, mostly pretty low." 2600 you say? Hmmm. That should set off a few alarms.
Say hello to my little sig.
I think it's below acceptable unless you are really on the ball and have a lot of things automated and properly organized and little to no emergency troubleshooting to deal with. At the community college I attend there is one Network Admin and 2 help desk goons. Admin *only* deals with network-related problems and basic telephone system issues. Helpdesk just helps out with end-user workstation issues. They seem to get along well enough, but the admin has lots of old software and hardware and an old PBX to deal with, and desperately needs a qualified assistant. Im working as his assistant with the work-study program but Im not experienced or knowledgeable enough to help the way he really needs it. I can deal with maybe 30% or 40% of what is on his "to do" list considering the limited access and experienced I have.
Now, he said at another school years ago he managed 2500 PCs by himself with a handful of work-study students, but my understanding is that the network was in better shape than where he is presently.
By and large, language is a tool for concealing the truth. -- George Carlin
pi * 10 ^ -6
366:1 includes management and combined positions. Public sector (at least city, county, school, state) tends to have more systems and less techs. My shop is a public sector shop.
Our system/facility counts:
~1100 desktops/laptops
~45 network switches
~10 physical servers
~10 virtual servers
4 storage systems
1 managed wireless system
7 schools, 4 admin buildings, ~2300 students, ~500 teachers/staff
Our personnel:
1 manager/DBA/server admin/network admin (combined position)
2 FTE tech support on staff now
1 open tech support position (open since September, and open from Feb-August prior to that)
Our biggest problem is we can't pay nearly what a corporate environment can offer, even though there tends to be far more individual responsibility in our positions. If I chose to work in a corporate environment, I could easily be making twice my current salary with my experience, education, and knowledge. But then I'd be bound by constraints the education field doesn't have.
The job situation in our area is also the reverse of the rest of the U.S., if you want a job (even one that can pay 6 figures), you just need to have a pulse, pass a drug test, and be willing to work on an oil drilling rig. They even have tech related jobs that pay in that range.
We run a software/hardware management system and centralized imaging system. It saves us a lot of time/hassle, but still can't take the place of one or two people, especially a higher-level server/network admin or DBA.
When we compare to other schools our size and larger in our state, those schools tend to have a better ratio of systems:techs.
We have 9 people to take care of around 3000 users with over 11,000 computers. We are a school system of 17 high schools and 5 offsite locations. At one point we were down to four people. Our schools are a state school system so they are spread across the state. The 9 of us do it all. Network, servers, computer support, etc. I do most of the switch work since I am the "Cisco guy" of our crew. So those of you who think your struggling with a hundred users or computers, count yourselves lucky! SMS, RDP and VNC = three sanity saving tools!
According to the Bureau of Labor Statistics, 565,700 out of approximately 145 million jobs in the United States last year were for computer support. That's one of every 256 jobs. However, a large portion of workers don't use computers for their jobs. I couldn't find statistics for that, but whatever the real ratio is, it's lower than 256:1. It's safe to say your 300:1 ratio is well above average.
This space intentionally left blank.
You're seriously understaffed, and it's normal.
> Are we seriously understaffed, or is this normal?
Yes, you are seriously understaffed, and yes, this is normal.
Like sibling, I don't envy you either. There's nothing worse than having to clean up after contractors, and then try and waste time teasing out WTF they did (because they obviously won't let you in on all of it, if for no other reason than to insure future contracts).
Quo usque tandem abutere, Nimbus, patientia nostra?
I read a couple of years ago from somthing study that Usenix did that the industry average of servers per admin was 60ish for Unix and a little 80ish for Windows. In the shops I have worked in, the average of *nix servers to admin has usually been between 40 and 50. The average of Windows server per admin has been closer to 80-90. But that's because those shops simply had more Windows servers for nominal functions (DCs, wins servers, etc.) The numbers of users per server directly served on the *nix side was typically higher because you had fewer backend process type servers for the Unix side than you did for the Windows side.
my personal experience as of late is this: our company, before downsizing, was around 100 employees in 3 locations. 2 exchange servers, multiple AD servers, FTP, etc. roughly 15 servers all windows. with that many desktops and that many servers i could handle it all. once you whip the servers into shape and get some desktop management going it's smooth sailing. a good IT admin with experience can handle quite a bit more than someone with say 3-4 yrs experience. i can handle 200 employees and 15 servers if they are all in the same location
I worked for a major Novell distributor in the early nineties. At the time, the SUGGESTION from Microsoft for help desk staffing was 1 tech per 50 - 100 users, depending on the level of automation that could be done (this was Win 3.1, including server). Novell recommended 1 CNA and 1 CNE per 30 - 50 servers, but that varied according to the applications running on the NW servers.
The reason Novell bought USL (Unix Systems Labs) from AT&T was that 1 Unix admin could support as many as 50 - 75 servers, again depending on application type. UnixWare could run Oracle DB, and that meant there was a requirement for support admins for those "specialized" applications.
In 1994, Novell spun the Caldera group (which started as a lightweight desktop replacement for Windows) off from Novell. At the time, I was still closely connected to Novell's Unix products group mostly through Kent Prows, who had fathered UnixWare through development. I was told at the time that Novell saw the Caldera project's success, and immediately ceased further development on desktop UnixWare, because "this Linux thing can do everything UnixWare can." That was from one of my other contacts at Novell.
The reason for the digression was that Microsoft had popularized Windows through the CIO and IT departments as "minimum wage administration" (verbatim from an MS distribution rep) and Novell had the burden of showing that UnixWare had all of the advantages of Windows in ease of use, etc. However, they had to get over the fact that Corporate Types have an intrinsic desire to build an empire, and that meant plenty of foot soldiers, and hence the bigger Table of Organization meant more pay. They (CxO's) like that Windows takes more staff; more staff, more pay.
BTW, at the time, Macs took 1 admin per 200 - 300 workstations, because of the better quality of software, and more necessity of reliable OS; because Apple had to support all of the Macs out there in home user land. I don't think that has changed significantly.
So there you have it; because I sold Unix systems to all levels of the Federal Government (the CIA loved SPARC stations; but they ground them up when they were obsolete), I had to be pretty aware of these numbers.
You can probably find out today's staffing levels from the respective OS manufacturers, search their sites for "enterprise staffing levels".
And good luck with your quest.
Dave Lawson
dot-sig.
I have been doing this IT (for this discussion I define IT as server and end user support, not programing) thing for quite a while now (15+ years).
My current gig will be seven years in a few months and what a roller coast it has been! After some cost exercising that was completed in Q1, there is only three of us in IT left and I was made the head cheese. We already ran a flat team where admins have to perform user support as well as their server stuff, so that did not change and in fact is the only way we have survived so far.
Stats: Three IT to about 375 users with computers with about 75% being laptop users that DO travel. We are currently sitting on about 150 servers with 90% of those being VMware virtual machines. Other fun stuff includes two phone systems (on 10+ year old PBX and one 3Com "IP" based system), 60 BlackBerries, we also manage all aspects of the three cell phone carriers our company uses(oh, that means about 40+ cell phones we support), tons of printers, one AS/400, and much much more!
How have we survived? Everything is in ActiveDirectory, making just about everything a virtual machine, picking IT tools that actually help (Track-IT, Rove's MobileAdmin, VMware VirtualCenter, TeamViewer), out sourced the after hours call support to a group that calls whoever is on-call and takes the initial notes, but no trouble shooting(fear not this call center is based in the USA!), keeping everyone involved in everything (cross training is GOOD for your team!)
Having social skills! Everyone (including me) on the team is not a StarTrek/StarWars geek and can have a meaningful conversation with the end users. We work very hard to have a open communication channel with upper-managment and decision makers, so that they know when it is time to buy us new shiny toys so they do not have to add anyone to the pay-roll. You would be amazed at how quickly things go your way after you butter up the VP's admin assistant! "Oh I love what you have done to your hair today! Can you get Mr. Man to sign this purchase approval for me?"
We are also very firm with our users, when they catch us in the hall on our way to one issue we say, "I need you to submit a ticket for what you are asking for and I will come back in a bit." Many do not like it, but we always spend the time to explain why it must be done. Only VP's get around the ticket requests, but hey you gotta scratch their back if you want their support when you need it.
The answers are going to vary wildly based on skill of the users, corporate environment and what they use the computers for. I've seen an insurance company that had 1 SA per 500 machines. They were all running a locked down version of XP, they were almost a kiosk. The users had 1, maybe 2 programs that they used, the insurance system and a mail client. It worked out well. Think blue collar behind a terminal.
On the other hand, an office full of 'empowered' users each with admin rights on their own vista machine, the need for using many various programs and various duties, you may see as low as 50 to 1... Executives constantly on the cutting edge...
Do not meddle in the affairs of sysadmins, for they are subtle, and quick to anger.
http://adsg.syix.com/linda/tokenring.htm
I use to work for a K-12 school district. 1100+ students, 100+ staff, way over 1200 devices (600+ laptops along with printers, desktops, projectors, etc). For 7 years I was the one and only techie. And they wondered why I couldn't keep up with the workload! Of course, having a number of 8 year old desktops didn't help.
I finally left.
But I've heard they now have 3 1/2 techs, and they still can't keep up. :)
Sounds like you are inept because 120 employees/users spread out in 4 offices should be simple to manage and support from an IT perspective. I have supported 100 users between 2 offices and after the first month of almost daily help calls from multiple users, I revamped the IT infrastructure starting with the servers and reduced help calls from 50 per week to less than 5 per week; of those 5 at least 60% was due to IP address lease issues which were controlled by corporate policy. Oh, and I was a support staff of 1 plus a manager.
So much is dependent on the specific details of the computing environment. How many servers per day can your administrators give the requisite level of attention, each and every day?
Clusters are easier to administer than stand-alone servers. Production servers can be easier to administer if you enjoy a proper development environment and change management processes, or more difficult than the development environment if you lack them. What sort of security risk profile must admins contend with? Are there other administrators? How well are your efforts coordinated? Are end users mucking about with the system? Do any require administrative access? How finicky are the applications being served? How robust and reliable is your organization's non-systems infrastructure? What's the hardware budget? The tolerance for process failures? What are the organization's expectations? How competent and responsive is the management? How homogenous is the infrastructure with respect to hardware, operating systems, and applications?
Ever watch Star Trek? Define a "level 4" scan process for your servers with a focus on manually verifying it's complete functionality and the basic sort of checks and audits that can be performed with the system still operational, log review and process audits and such. How much time per server will it take an admin to complete each task? If an admin can't get this done for all of her servers in one day and still have time for lunch, her hands are too full.
We are an educational site that runs around 1000 PCs running Windows. We have around 200+ applications installed on those PCs (each and every one), with about 2700 potential users. Our Windows machines are totally automated from cradle to grave. There are application install and remove scripts for everything we do. The build is automated, the rebuild is automated and can be done remotely. The problem is NOT managing the machines, the OS, or even the users. The problem is managing the applications.
Our automation is setup so that we could easily scale to 10,000s of PCs. Putting a PC on a users desk with Windows OS, and patching it every month is a no brainer. Our PCs run in lock down mode where the users run without Admin privs. The users have roaming profiles and folder redirection with our AD setup. The users can easily move from machine to machine. We have around 25 servers to manage our desktop machines.
The reason I say that applications are the issue is because for every application you install, you must manage that application over its life cycle. This includes things like making sure the machine has the neccessary pre-reqs, making sure licensing is taken care of, making sure that software dependencies are met, etc. Removing apps can also be painful if you don't take care to remember how one vendors app integrates with another. To top it all off, the software vendors for Windows are a very nasty bunch. Every vendor has a different install method. From InstallShield, to MS Installer MSIs, to NullSoft, to hand written VB installers, to BATCH files, etc. Even after you get the app installed as Admin, many vendors applications won't work as normal users becuase they don't yet know how to setup a new user when a user logs on the first time. Many times a vendor will program their application to use the "All Users" settings, other times, they will use the users own personal profile. Software vendors are a royal pain-in-the-ass. Also some apps require patching as well as security updates.
Most of the problem was/is caused by Microsofts inadequate attempts at actually "defining" a working user environment. Half of the software vendors out there just don't understand exactly how a non-admin user should interact with the system. Some work with roaming profiles correctly, others have never heard of roaming profiles or folder redirection. We get some of our applications out of the back of college text books, written by grad students at other univeristies. Those grad students just don't understand things like gettting your dot-net app installed on a users box, especially if the user doesn't run with admin privs. The Windows user environment is very hostile to IT administrators.
On top of that are the issues caused by anti-virus. You take a nice machine running multi-core at 3 GHz, then you stuff an anti-virus package, and that machine will run like it was 4 years old.
Our ratio is 3 guru admins/architects for those 1000 machines. We also have 1 'nix admin, 1 Oracle DBA, as well as a 1 full time hardware guy. So 6 full timers (in the back-end), plus some student TAs managed by 1 full time help desk operator. We are most of the time pulling our hair out. We've been "planning" on moving to Win7, but with our monthy pace of patching and adding/removing applications we just don't have the time for a new architecture. Having to manage bringing everyone on to the new architecture at the same time with the migration of 200+ apps just IS NOT GOING TO WORK. So there would be some time where we are running both the old and the new system. Quite a daunting task given the number of admins we have.
The other problem is one of finding the right application installers. Many people just don't get why they should take so much care when installing a new app. When your one application rollout "breaks" the entire system you will play not 52 card pickup, but 1000 machine pickup. That is an ugly situation to be in. So most of the time we sit around and run testing scenarios to
My outfit is a call center, We have 20 or so servers and 150 computers and 300 users.
we have 1 dba/asp programer and 1 sys admin/helpdesk (me)
The statistics are poor. We're in a mixed-use environment with mostly Windows for users and mostly Unix variants for servers, including HP-UX, BSD, Solaris, and Linux. The best data I have ever been able to come up with is one tech per 100 units. I've never counted routers, switches, hubs, and wiring in this, though I think you could make a good case to add them in. Some of those are a lot more onerous to configure than a garden variety PC. One thing that helped us was a standard-build PC. Store all data on backed up dual servers so if a PC breaks, you can replace it with an 'identical' PC easily and quickly. We kinda screwed up originally because we were IP rich with eleven Class C networks, so we used IPs to identify PCs and hard coded them, and also used one Class C per building, which was a big waste. It was a bit of a challenge to move to DHCP allocation when our Class Cs began to fill up, but we managed to avoid a lot of subnetting for a couple of decades. You probably couldn't get away with that these days.
How about a moderation of -1 pedantic.
Simple thumbrule number I use is 70.
But that should account for enyone and everyone that is on the payroll for IT, regardless of what they do (management, or sneakers) and calculated as a "full time equivalent", that is two half-time employees account as one full time.
The figure 70 is then adjusted for:
1. size: larger to company adjust upwards
2. centralisation: dispersed company adjust downwards
3. IT Maturity: more mature, adjust upwards
4. other...
Any adjustment should be kept with 10 units (so 70 +/-10)
In my experience you end up with a figure close to 70 anyways, and that figure is pretty much spot on with any company out there.
Robert
IT Management consultant, Stockholm
Where I work (a large printing/office products company in the Eastern US) we have TWO dedicated IT personnel.
We have 800 employees, 600 of which we are directly responsible for supporting. That's it. One IT manager, one Systems Engineer (me).
Corporatism != Free Market
Your boss, and offshore.
I work in an office with 400 people. There are 3 local admins, but we are not allowed to use them. Their sole job is to keep the phones and internet service going.
If we have PC or Server issues, we have to call our company's global help desk in India to get a solution.
And how much of it is automated and remotely accessible.
For example, our IT department has 15 people:
1 Director
1 Manager of Systems Admin
1 Manager of App Devel
1 WAN Specialist
1 Programmer
1 Helpdesk
4 High School Techs (each looks after 2 or 3 schools)
2 Elem School Techs (each looks after 15 schools)
1 Electrician
1 Hardware Tech
1 Videoconferencing Tech
We support over 5000 computers, over 100 servers, in about 70 buildings. 13,000 student accounts, and 1,600 staff accounts (at the district level, another ~1,600 at the school level).
Between the WAN, programmer, and helpdesk, we remotely support 90% of the computers and all of the servers, since just about all of the servers are Linux or FreeBSD, and almost all of the computers are diskless Linux stations booting off the network. Pretty much everything is automated, and the individual computers have been reduced to appliances (no harddrive, no cd-rom, all onboard video/sound). When one breaks, we just swap in a new one, update the DHCP config, and they're off. Less than 5 minute turn around.
We're actually looking at expanding the helpdesk and videoconf side of things, as the school techs are having less and less to do.
If you can centralise the management, you can do with fewer staff. If you have to touch each individual workstation for updates or installs, then you will need more staff.
Our IT dept here where I work is 1 Systems analyst, 1 network admin, and our boss. That is for about 200 pc users across 27 locations, not big, but our ratio is what works for us, so we don't rock the boat... maybe if our users were a little more savvy we could run with 2 people, they tried that and it wasn't enough, but 4 would be too many. Really this is one of those hard to predict things that could be 1 for every 10 people, or 1 for every 50 people. In our case, we try to automate as much as possible, eliminating a lot of the mundane things that are just annoying to do... little things like that help keep everything balanced and maintain a reasonable tech to user ratio.
And often the old problems are replaced with a whole new batch, which might even be worse than the first set.
Also, killing people is generally frowned upon around these parts.
Yes, the BOFH is my hero.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
We had two admins for 160 client machines (labs included), 12 servers and fifty users.
Feel free to apply your own math to that...
I just don't get... eh, ugh... never mind. This post wasn't worth the research I put into it.
Its not a question of how many, but whats best. IF you have systems designed right, you could get away with a fraction of what many places have if the sites are localized enough. In education IT, our district has 7 techs (this is including the network manager) for roughly 1800 machines and 6000 users and realistically we could use twice the number of techs unless we got enough of a influx of money to completely redesign our districts IT structure to work better (its education so there is fat chance of that happening). The district next to ours has 2 for a little over 900. Then you can have 3-4 for 500 which i see often in small business.
"Slashdot, where telling the truth is overrated but lying is insightful."
Shouldn't it be how many users/computers per administrator?
Respect the Constitution
We are a small business and every user is her own admin, so lucky to have a personal admin !
30 users, 2.5 techs. One does half tech and travel scheduling. All Macs, all two years older or newer.
One per 70? Perhaps if you are running a ton of clone servers or heaven forbid working at an ISP handling boxes for customers.
In the *real* world one normal admin per 35 or so is considered more normal for unix staff (with a senior admin for every 3-4 normal admins). While I have seen the numbers quoted by the parent they were very for generic servers in function and basically tons of copy (virtual) boxes. In addition if you are running a good deal of unix servers you more than likely have a lot of different functions spread out among clusters, one person responsible for all these apps as well as keeping the boxes well maintained, troubleshooting upgrades, etc... can be quite tiring. Unix boxes these days tend to do significantly more than they used to. On the flip side, managing unix boxes these days is quite easy thanks to ssh/cfengine/puppet other automation.
When I first started doing unix for pay, mid 90s, we had one admin per two servers which was a mirror of the typical university environment where there were single admins for important machines and tons of lackeys for all the sun/hp/dec desktops. By the year 2000 this had grown to around 20 servers for each admin, with senior staff expected to be able to handle functions across any cluster. These days if it is a smaller company, there's one senior admin, who may or may not be the Network/IT manager and they typically handle 30-40 servers if there are a number of functions spread out. Of course having tons of similar boxes makes this number much higher. The most I've ever handled was 165 servers spread across two admins, three sites, four major applications, tons of clusters, including DB/Application/LB/Development/etc...
Also, if you have 70 unix servers and only one admin, the point of failure for your organization is beyond catastrophic. That poor lil' admin guy is going to get many of the "what happens when a bus hits you" questions from 3rd parties/partners/other departments.
--- I do not moderate.
I'm the only IT person for a business with 130 users and 10 locations plus a second business with 25 employees and one location. I'm responsible for everything - desktops/laptop support, phones and blackberry, routers, servers (file shares, exchange, DCs, sql, citrix, term server, etc), and purchasing. I'm also on-call. Honestly, some days I'm overwhelmed and others are quiet. I have an IT consulting group on retainer for when I need advise or an extra person for a day. I also have a very supportive non-IT manager.
The truth is that there is no benchmark for this. I am a consultant and tend to take a sysadmin role for clients.
If you use Active Directory and store user files on the network then you can do stock images for each model of machine and a broken unit is a 20minute re-image (or swap fresh machine in from the pool) and your up and going, documents and all. This is where a directory services' up-front costs become justified.
With Active Directory I can manage machines and users very efficiently, keep user's files safe with shadow copies, backups, etc, and deploy software and printers to users easily. For linux to Active directory look at likeise-open or centrify for single sign on with the latery able to do group policy on linux machines.
I have 4 techs and myself. between the 5 of us we handle about 2500 or so users across our clients. Our clients that have been with us for a year or two are all have some sort of AD setup and have a much lower computer expense than before. oddly enough, newer clients account for larger expenses because they havent standardized their computers ad require more trips to their site and more billable hours.
I would imagine that if I had only established users, with computers on AD then my crew could handle 3000-3500 users without much overtime. If we did no managed computers then I think that 250 users per tech would be pushing it.
If you just compare those numbers, 600 vs 250 you can pretty easily see the cost savings for a managed network, either through AD, network, or other LDAP. a 1200 user network could be reasonably run by 2 IT guys vs really needing 4 or 5 to do the same job otherwise. let be conservative and say 4 guys at $40,000 each vs 2 guys at $50,000 (higher skills for 2 techs vs 4) and you see a $60,000 gap, which is much more than the CALS and servers needed for 1200 users. 1200 users is still in the 2 ADDS servers arena. lets say $3500 per licensed server and $35 per CAL and you save money on year 1, next 4 years are free!.
Right now it is kind of handy because my guys work their ~40 hours doing stuff on managed networks and pull 'overtime' going to customers sites or doing old computer triage and repair and get paid part of the service fee.
If you are at 600+ users per tech, then you really should be on some sort of directory service like AD. If you are not, I suspect you are spending a lot of your labor dollars spinning tires and not helping clients/users very well. That equals more compaints, less praise, and likely a lack of raise or bonus.
The reason there are no good rules of thumbs is because environments and applications vary widely. Are the applications canned or custom? Are we supporting general services like file sharing or printing? If so, are we talking about printing documents here and there, high volume legal or medical printing, floor plans on a giant honking plotter, pre-press ad copy? Are the files the usual office flotsam and jetsam, high value CAD files that demand rigorous snapshots and backups, raw video files that require high speed workstation storage? Are there natural downtime windows to perform maintenance? Are there internal or external SLAs to meet? Are users given administration privileges, or is everything centrally managed? What mobile devices are in use? Are networking needs managed in house or contracted? Is there need for VPN access? Is mail storage done in-house or outsources? If so, what's the spam and virus filtering? Are there quotas? Synchronization with mobile platforms? Integration with an instant messaging platform? What about hardware? Does the hardware cycle match the warranty cycle, or does the company expect the IT staff to do their own maintenance?
In other words, the correct ration of admins to users or computers is "as many it takes".
Seriously. What does your experience of management decisions teach you about probables like that? Be glad you have a job, and never, ever, under any circumstances, dare to even THINK about unionizing.
But if you do, I recommend the Teamsters.
``Tension, apprehension & dissension have begun!'' - Duffy Wyg&, in Alfred Bester's _The Demolished Man_
Since your question was rather vague, I've got a few questions for you:
- Let's start with your definition of Admin, Technician, or "Support Tech".
- What responsibilities do they have within the environmnet?
- Are they simply Desktop/End User Support, or do their roles expand beyond that?
- Are you lumping all support people into this same category?
Assuming you're talking about strictly Deskop / End User Support (aka. the guys who show up at your desk to fix a problem), the answer depends alot on their skill sets, the complexity of your environment, their responsibilities, level of automation, and a whole series of factors. The Help Desk Institute (http://thinkhdi.com) has a lot of information that can help. Given that you are a Windows PC shop, Microsoft has a bunch of analysis tools that can help you detirmine the complexity of the environment and suggest ways to optimize it. (Google: Microsoft Infrastructure Optimization model (IOI)).
Assuming you're talking about an entire systems support organization, I've seen IT consume up to 60% of a company's total staff (including developers, project managers, and systems analysts). Within that, I've seen up to 30% of those people to be "systems support" (Service Desk, Desktop Support, Data Center Operations, Telcom, Networking, Windows/Unix Server Support, DBAs, and various Architecture and Engineering teams).
Since you're probably looking for a swag, here's what I'd say:
For 900 seats and 9-5 operations...
3 persons answering phones
5 persons visiting desks
3 persons managing servers/storage/mail (add 2 more for networking/telcom)
3 persons managing patching and making everyone else's jobs easier (aka, engineering/projects/packaging/scripting/imaging/etc)
If you want to know where my math comes from, just ask. But it breaks down on a few assumptions of having a diverse team with varying degrees of experience and skill sets. Of course, if you have a handfull of rockstars, the numbers change, but in the end, the salaries will probably add up to being the same.
Clearly this ratio can be impacted by many factors and while there is no true standard, I found this video calcuation method reasonably accurate: http://news.zdnet.com/2422-13569_22-155252.html
Anyway it took a whole bunch of work, we inventoried all the hardware, and made groups of 20 that were identical. Then we said NO to any application that broke the current applications in the slightest. We said no when they wanted to move to the area by the window, where all the misfit machines sat on the desks. (All the rare parts in one unused pod).. It became required to use it when there was no other place to use.. By that point we were down to 1 crew of 7, who were having problems trying to look busy enough not to be downsized. /hardware remotely, (the manual busy work she (our boss) gave our crew too often).
After getting in new boxes with matching hardware, we were down to four people who were less busy than the security guards. One of the crew built an impressive counterstrike map of the building. Our boss got awesome at quake. I practiced coding, and built scripts to inventory all of the software
Anyway for 900 boxes 3 could be plenty, or horribly understaffed.
cheers
I have. We had two SAs who handled everything on that rollout. I was one of the two. We installed and maintained all software (that was written and pushed to a central server by a different staff). We installed and maintained all equipment, both field and server room. The standard setup was each group was 10 to 15 people. One secretary used a desktop and everybody else had a laptop. All laptop users got an individual laser printer for home use. Each group had two impact printers for forms and one networked laser. Each group connected to their own database server. There were three additional servers that controlled other functions. Every single machine ran SCO OSR 3-point-something. Hardware was IBM for everything.
Us 2 SAs did everything. It didn't matter how small or large the problem, from replacing a USB-to-ethernet dongle to rebuidling a server (I could fully build a server, pulling spare hardware, imaging from tape, and restoring all databases from online backup in under two hours. We had that shit wired, I tell ya!) to taking calls from people who didn't know how to turn off the reveal codes in WordPerfect, we did it all.
Everything that could be scripted was. Our morning checklists and reports took a half hour, tops. On a good day, that was all we had to do all day. On the worst of days, we might work hard. But bad days were rare. We could take our vacations and know that no matter what shit hit the fan, the one of us who was on-site could handle it. We had the wonderful luxury of being able to walk around the user groups and ask people if they needed anything. They almost never did.
Our total user base was about 300 people. So I'd say if things are designed right, 2 people can handle 300 easily.
Of course, there were 25 or so admins and desktop people on the Windows side of the house, taking care of about 1200 users. They ran around looking like they were doing important stuff all the time. And I guess they were. Their stuff broke so much, they were constantly being rewarded for rescuing some project from the jaws of disaster or fixing some irritating problem that had plagued their users for years. Those poor sods hid in their cubicles most of the time; they didn't dare walk among the user population for fear of someone throwing something at them or, at minimum, being constantly harrassed by users pleading "Could you take a quick look at this?"
Our users just did their jobs, working on hardware and software that just worked, reliable as gravity (well, nearly) with no drama at all.
You can see what's coming, can't you?
The higher-ups started wondering aloud why those two SAs over in the corner never seemed to be running around in a panic fixing things. "Don't they have any work to do?" The higher-up attitude toward the Windows guys was completely different. Hell, I remember one of them getting an award for recovering data from a crashed server. They actually rewarded the guy with a certificate and a little ceremony because he had backups, something we took for granted in our little world.
Obviously, it couldn't last. All our apps got re-written to Windows. All the Unix stuff got ashcanned. Our user population got folded in with everyone else and forced to use the standard Windows-image machines.
And we now run around putting out fires with no time to catch our breath.
Man, those were the days. 1 to 150 was a breeze. Nowadays, deskside support is at about the same ratio and we're always on the verge of burnout, always working harder, always falling a little further behind. As much as I love my work (and I do, dearly, love helping alleviate the pain of a user who can't get their work done until I fix something), I'm *seriously* looking forward to retirement.
1250 PCs, 10 ppl in tech support, 2 in "servers + network" support, 1 servers and 1 network supervisor
I'm the only "IT" guy, we have 5 others (1 manager/programmer, the others programmers) that are upgrading/developing internal applications. I take care of all the support issues (except for those internally developed apps). We've recently hired a sort of level-1 person, that is here part time. Job demand is such that we can just barely jusify having the part-timer - as this person's position has to be "billed out" to the other departments. Myself, I'm not billable for most of them. Strange, I know. Overall, we have around 200 users, according to active director. We have around 20 servers. Computers, we have around 300 or so. Why, you might ask? We have computer labs that the public can use...that we have to maintain. By myself, I get more and mroe behind, projects get delayed, ideas are never fleshed out. With the part-timer, I get enough breathing room that I can do more, even though I have to walk this more technical person through things...at least they get it, where our users don't.
42
Windows is not the answer.
Windows is the question.
The answer is "NO."
I was the network manager for a bank with 17 branches, and 3 office locations, about 200 or so employees, and about 40 servers (production, training, and test).
Total IT staff at the bank: 1 VP of network services (I reported to him), 1 network services manager (me), 2 help-desk/branch support guys (reported to me), 1 DBA, 1 core banking application support/administrator, 1 IT dept secretary (actually VP's secretary, but she helped us all out), and 2 operators to run the check processing/statement printing systems.
Now I am the IT director for a small private school. We have one school and one office/tutoring center, about 70 employees and 128 students, 20 servers, and a bit over 200 laptops and desktops, as well as other stuff (routers, switches, firewalls, remote access, VOIP telephones, printers, copiers, IP cameras, AV gear....etc)
Total IT staff at the school: 1 IT director, and two technology teachers. The teachers are in the classroom most of the day, but they help students and staff with the "help-desk" stuff, and escalate severe problems to me. Administrators come directly to me for support.
-ted
I can remember one of Support resigning in ecstacy. "I'll be in charge of 13 workstations! THIRTEEN!!!!"
Not sure how that worked out but I think we can safely say that the range is considerable.
Maybe you compare number of "support tickets"?
Are you still using windows xps? This is a support hell. Do you have something called network?
Feeling stagnation? Install Windows 7, he he he.
Facebook: is not someone to copy. The security of their systems is highly suspect. While I don't have any first hand knowledge about facebook, I know that many, including the most popular, 3rd party flash games are compromised. Ever noticed how many unplanned outages they have?
Today, any company that ends up in the news for network or security related items is **not** a company I want to:
a) do business with
b) invest my savings with
c) encourage everyone I know to use.
Security matters. Do any of you still shop at TJ-Maxx or K-Mart?
While this is a different situation from the article, it indicates that the number of people required is to a large extent determined by the degree of automation and the discipline of rigorously implementing that automation. Without configuration management, life is much harder.
County government: 300 users. Two IT Specialists and myself. Mix of Unix, Linux, but mostly MS.
This would be a great topic to create a simple survey of. Such as 1-25 machines per admin, 26-50, 51-100, 101-150, etc. Pure and simple, regardless of workload. How many machines are managed by how many administrators.
I've never done IT in an "IT shop", whatever that means, exactly. But my experience is 1) in a larger national medical service provider, and 2) a small regional/rural hospital.
1) The national provider had approximately 250 Linux "cache" servers located remotely, about 3,000 XP workstations (with many more than 3,000 users), and approximately 25 national campus servers. The support was tiered; I was on 3rd tier with 2 others, with approximately 15 people on 1st tier and 5 on 2nd tier. I spent about half my day on support calls or fixing workstation/server problems (I was the lowest-level 3rd tier) and the rest were spent doing equipment deployments and requisitioning. In addition, there were 2 guys doing network/host security, someone who did all the Windowsy admin work, two network (Cisco) guys, a user requisition gal, and two database guys. (There were also about 8 or so guys doing the development for internal software packages.)
2) I was, essentially, the sole supporter of workstations and servers (there were two other functional staff, one doing biomed stuff and printers, the other doing support for the large monolithic terminal app everyone used.) There were 250 workstations and about a dozen servers. The servers were not maintained well in any sense of the imagination - they were spread out throughout the facility, and were a real headache. The workstations were mostly OEM XP installs with crapware installed, et cetera. Realistically, there should have been another person in my role (minimum), and we should have worked towards a better arrangement than what we had, as there simply was not enough time for me to do it on my own after triage.
So, basically: it depends. I'd say the reasonable number of support people decreases as the organization grows beyond a certain point; likewise, the number of justifiable support people increases as the organization grows from 1 up until a certain point.
A well-run shop is not going to need as many people as a poorly run shop. But, the size of the shop often dictates whether there are enough competent people to make it run smoothly, so it's a bit of a catch-22.
Case in point... I did support on a small Mac based network for a while. It was just me, and 9 users with some fairly specific needs. They were just at the point where they actually needed someone at all, but it was enough to be a FT role. After a couple months, i'd gotten rid of most of their issues and managed to get things running smoothly and trouble-free. My reward was the closing of my position.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
I ran a whole IT department by myself with hundreds of servers. Life was great. I used tools that I understood, was confortable with, were tested over time. Everything ran smoothly. Then new management came in, changed everything, now it takes > 4 admins for twice the number of servers, and we're always behind with tickets and our support is worse. So my answer is "it depends".
I've used a tool for a decade now that I really like. It was developed by IBM and MIT to determine staffing levels. Basically, it asks you to go through your entire organizational and IT structure to answer questions, weighs these and provides an answer about how many FTEs are required. It does not break this down into programmers vs. trainers vs. help desk. The project was call Project Athena. You can find the write-up on DocStock or InternetArchive (I don't recall which). It was a tool used by the State of Michigan as little as four years ago but has disappear from their sites. Examples of the types of questions follow: 1. Number of computers (workstations, not servers) in your organization. This is divided by 500 to give a ratio. 2. Number of server-based major applications in terms of FTE requirements. For instance, you have a heavy email system that requires someone taking care of it half-time. This would contribute .5 FTE to the count.
3. Number of user applications, for example 500 Microsoft office, 10 Visual Studio Enterprise. These are added up and appropriate ratios are applied.
4. Intensity of use, # of employees using their system between 50% and 100% of the day get 1 point, occassional users get .25 points. Again these are added up and ratios are applied.
5. Other responsibilities, phones, copy machines, etc.
I monkey around with the figures trying to keep them meaningful and usually end up getting what I need (except possibly in my current job) in terms of human resources. The system is quite comprehensive and flexible. Even though it was designed for educational institutions, I believe that it's generic enough to apply to business.
I managed DNS service for a region (many countries), It was only 8 machines. It was thousands of users (8000, perhaps more), there were 3 of us (not exclusively doing this, but I mention it this way for contrasting purposes).
In another job I had only 8 users, and 16 machines, but the software was highly specialized, that was more time consuming than the example above.
So are you understaffed? You are the only one that has a chance at knowing this.
Are you constantly staying late and working out of hours? Then you are understaffed, disorganized or both It is that simple really.
I have been in shops where they expected exponential growth could be managed without extra personnel, you should learn to identify those situation and plan accordingly (either reduce growth if you can't hire more people, or hire more people).
IANAL but write like a drunk one.
If everyone has the same desktop, keeps their files on a central server, and you can rebuild the desktop with a single command, then the correct answer is likely "one"
Reality strikes.
Not everyone will have the same desktop. Only a very few copies of AccPac accounting need to be installed.
Someone will need photoshop. Or InDesign. Someone will need Autocad.
If you can take the time to automate every install. Every computer is associated with a profile to install the right stuff. Then life is easy.
I've had 4 sysadmin jobs.
Space Physics -- 10 NeXTs, 2 RS6000, Stardent Titan, Dec Ultrix, Myrias SPS 3, couple PC's. Easy.
Math Dept -- 3 versions of SGI IRIX , 2 versions of Solaris, 2 dists of Linux, HP HPUX, RS 6000, 5 versions of Microsoft. 250 machines total. That kept me hopping. About half my time was spent on the 30 machiens that ran some form of windows.
YottaYotta. All linux. Had time to make automated reinstall systems for the developers to crash regularly. Easy.
High School. FreeBSD servers, windows 2000 clients. 60 machines total. Windows took up 90% of my time.
Generally:
Managing a hoard of anything isn't much harder than managing three of them.
If you have an OS where applications can be completely divorced from the OS (*ix) and can be run from a network drive, then almost all of your individual workstation customization is trivial. (symlink to the application for light weight stuff -- rsync nightly for heavy weight apps)
If you have an OS that is smart enough to recognize hardware changes, and not bork on you when a disk image is moved to a slightly different hardware platform (Linux, *BSD) then you are golden. You don't have to manage the combinatoric explosion of N different motherboards combined with M different desktops.
If you have a setup where users cannot install executable software, the crisis count goes way down.
In short: Managing a hoard of machines is easy if you don't have windows.
Third Career: Tree Farmer Second Career: Computer Geek First Career: Teacher, Outdoor Instructor, Photographer.
This is really a management decision.
How many admins you have should be weighed by the department manager. That is decided by the workload. How long does it take to get issues resolved?
I ran a shop with about 150 Linux servers, and about 2 dozen Windows machines. It took 2 to 3 people to do the job effectively. The Linux machines ran themselves very well. The Windows machines did ok, but required manual intervention on a fairly regular basis. If you're not geared up for automatic work, then your numbers would go up.
Do tasks get completed in a timely fashion? Then the staff load is correct. If the work queue is too long, then you need more people. "too long" is decided by the business needs. Is it acceptable for a workstation to be down for 10 minutes waiting for a tech to get to it? Then you're fine. If it's acceptable for a workstation to be down for days (the staff can move to another one), then you're still fine. If the problems are not resolved fast enough (desk workers not able to accomplish their tasks on a regular basis due to IT reasons), then you have a problem.
If the staff works 40 hours a week, and all goals are being accomplished, then they are doing the expected workload. Management should have already accounted for fluctuations in the workload though. If you had 30 IT guys for 40 desktops, and the IT guys sit around with nothing to do all day, then obviously you're overstaffed.
If you think that the workload is too high for the IT staff to accomplish, that's something to take up with your manager.
Serious? Seriousness is well above my pay grade.
I used to be the only "IT guy" at a workplace of 110 Windows and OSX machines. If there was anything IT related ("build us a website","make this expensive new Windows-only printer we just bought work with that Linux thing") it fell upon me to do it. Of course, I was fired because I started to say "no" to their technically impossible requests leaving..... no one behind to do anything. Oddly enough, the company still hasn't got a new "IT guy" in after a year... but they did get in a 3rd party to replace their 14 year old NT4 server box that was sitting on the floor under some guy's desk. 110 users-to-0 admins is the same as infinity-to-one right?
50 users (about 20 of which are on the road) 10 windows servers (overprovisioned) 1 admin (me)
Large international finance company - trading, research, etc. Asia Office, ~2800 users, 3500 computers, large server farm, also backupsite which is also online and ready incase of disaster For JUST desktop support we had 10 techs (2 managers, 8 grunts) - the "helpdesk" only took down the details of who was reporting an issue, and we would respond. Desktop "team" also was in charge of all aspects of desktop support including maintaining software, setting up new hardware, and actual physical install/swap/moves. Large moves were done by outside contractors, then we'd just push the power button :)
There was also smaller teams for servers, messaging (mail and phone), networking and a very large inhouse developer group.
For what its worth we all kept fairly busy - not having to worry about servers was nice, but having each group seperated did cause issues when interests overlapped (ie. desktop team finds an issue with the network, but network team refuses to do anything because it doesn't affect them - ie. new fast desktops would boot before the ports came online due to port security settings on the switches).
Biggest factor is what kind of support is needed? Some of the admin people could go a day without email if something broke and only grumble a little about it, but on the other hand traders would expect instant results and would threaten you every second it takes for you to fix even the tiniest little issue.
You will generally need fewer IT staff per hundred users if:
- You have a large organisation (several thousand staff) and can use economies of scale.
- You have a homogeneous digital corporate environment.
- Your business does not change much from year to year.
- Your business is largely non-IT based; miners and dock workers generate less IT demand than office workers and engineers.
- Your IT infrastructure has been built up slowly and thoughtfully, with consideration given to interoperability, remote administration, and software automation and maintenance.
- Your IT department is allowed to devote resources towards improvement of service.
- Your IT department has extensive input on incoming upgrades, speccing and sourcing, and other uses of the IT budget.
- The IT staff is paid well and provided with an attractive working environment, leading to lower turnover and greater average ability/knowledge per IT person.
- The IT staff don't mind cross-training and providing a degree of flexibility and coverage for each other, which is a lot easier to make stick if people actually like working for your IT department.
- You have procedures in place for both reducing instances of and minimising effects of things like bad software rollouts, new interface implementations, and so forth.
- You have methods in place for allowing a sliding set of task assignments, so that if something big does unexpectedly crop up, people can be smoothly assigned to it in an actually effective way without their normal duties falling by the wayside, and it can be taken care of much more rapidly. You also have procedures for rapidly acquiring additional emergency IT staff who won't need three months' training to get up to speed.
So, yeah. There's a bunch of factors involved. A ten-person IT team might be needed for a fifty-person bleeding-edge programming shop built out of spit and Steve Jobs' old turtlenecks, but an identical ten-person team might be able to hold down a fifty-thousand-employee megacorp where the infrastructure is a locked-down 500-person HQ and two hundred sites specialising in physical labor and hosting a single indestructible greenscreen terminal each because 90% of their paperwork consists of the boss's clipboard and a work schedule stapled to the break room wall.
the management demanding it because that's what they want to beleive that everyone else uses
There fixed that for you. Either way, fixed or unfixed, the statement points out the problem: Windows is there because the software is not selected or even evaluated on technical merit.
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.