Softpanorama

Home Switchboard Unix Administration Red Hat TCP/IP Networks Neoliberalism Toxic Managers
May the source be with you, but remember the KISS principle ;-)

Number of Servers per Sysadmin

News Solaris Recommended Books Recommended Links E-books, Courses, Tutorials Articles Sysadmin Health Issues

Unix system adminitration is a job that can be performed only by those who have certain abilities. Like in art only a small number of people can reach to theo in this speciality. So ther are brilliant sysadmins, average sysadmins and poor sysadmins. Repectivly if we talk about any metric we are talking about average sysadmin.

The second factor is the number of sysadmins in the organization. I suspect that there is empirical "Softpanorama rule" which can be formulated as following: efficiency of sysadmins is proportional to the logarithm of the number of sysadmin in the organization. So after a certain number simply adding sysadmins does not work well.  That means that the heist productivity is probably observed in organizations were the total number of sysadmins is small. In this sense organizations like Google are doomed to be a failure despite high effort to recruit people with substantial talents. In other words they are often a graveyard of talents.

Efficiency of sysadmins is proportional to the logarithm of the number of sysadmin in the organization.

Sysadmin job is first of all job that requires the ability to tame the complexity of modern Unix and datacenter environment. Like one Solaris sysadmin guide aptly noted:

Winchester Mystery House [in San Jose, California] . . . was designed to baffle the evil spirits that haunted Sarah Winchester, eccentric heiress to the Winchester Arms fortune and mistress of the house. With 160 rooms and 2,000 doors, 13 bathrooms, 10,000 windows, 47 fireplaces, blind closets, secret passageways and 40 staircases, the house is so complex that even the owner and servants needed maps to find their way.

--AAA, California / Nevada TourBook, 1991

Sarah Winchester, listening to the advice of psychics, believed that if she kept adding rooms to the house, she would not die and be subject to the influences of spirits who had been killed with the Winchester rifles manufactured by her husband.

The UNIX operating system is much like the Winchester Mystery House without, we hope, the evil spirits. The original operating system has been continually enhanced and expanded. There are many ways to get about, and, like the owner and the servants in the Winchester house, system administrators frequently need a map to help them get from place to place.

To add to the complexity, there are many versions of the UNIX operating system based on either Berkeley (or BSD) UNIX or AT&T's System V.

There are a lot of discussion about what is the average number of servers per "average" administrator. Usually they are extemly superficial and do not take into account two major factors:

Additionally the answer depends on such factors as (see Number of Servers per Sysadmin):

If can also we approach this issue from the point of view of in-house costs vs. outsourced cost. If we assume that a fully outsourced server maintenance costs approximately $200 month or $2400 per year, then assuming $72K salary we will have 30 servers per sysadmin. At $400 per server per month it will be only 15.

In general question in its plain vanilla form is very misleading as another variable is the qualification and experience of the administrators). There are "super-administrators" that are capable to carry load of three or more "average" administrators. And they are "permanently entry level" sysadmins who do not know scripting and have limited ability to troubleshoot complex situations under high pressure (and such ability only parcially can be developed on the job; to a large extent it is a personality trait, like "courage under fire" in military). They are half-operators, half-sysadmin.

Also the load that one admin can carry depends on tools available. With well integrated and suitable for the environment  tools for software distribution, monitoring, log processing and configuration management that load can be at least twice lower then in "plain-vanilla" environment.

Even simple tool that allow replication of configuration like Kickstart/AutoYast/Jumpstart/Ignite/  (and/or Expect scripts) can substantially raise productivity of repetitive tasks. For example (RE [SAGE] Servers per SysAdmin):

Among additional requirements the key is support of several flavors of Unix. Supporting more then two cuts productivity in half or more :-(. Other worth mentioning differences might include:

Risk of doing something wrong is also higher for enterprise system administrators and they often need to follow specific procedures to minimize the risk of downtime due to patching of changes in configuration of the severs. that cuts the number of servers which one admin can service. It is especially high when datacenter is remote and does not have qualified personal on staff (isolated autonomous datacenter).

When you think about simplifying and partially automating administration in the large enterprise environment, then large and expensive systems like Tivoli, HP OpenView, Sun Management Center are naturally come to mind.  Some of them like OpenView are helpful, some like Tivoli are less so, but both require regular maintenance and support by themselves, diverting part of the resources from maintaining production servers. For large corporation this worth the trouble for small your mileage may vary.

Still but you should not forget about the mini-tools and open source solutions. such tools as screen, VNC, Teraterm, lftp, expert, Perl are great help. IMHO one of the greatest tool that simplify Unix administration is an orthodox file manager. If you do not use it, please take a look at OFMs. Windows top OFMs: FAR and Total Commander also can be extremely helpful in multiplatform environment and for transferring files from Windows desktop to the server and back. Unix OFMs are either more limited (deco) or not completely portable (Midnight Commander), but still Midnight Commander make a lot of sense for Unix administrator. The other useful and largely under appreciated tool is folding editor like THE or at least VIM 6 and later. I believe that for sysadmin tools one should stick to tools that use text format in configuration files. They are much more manageable that tools that use binary formats.

There are also pretty good open source system monitoring tools such as Nagios (it has important advantage that the packages are supported for enterprise version of Red Hat and Suse so installation is a breeze)

ssh and VNC are another two cross platform tools that can simplify many tasks without any additional maintenance efforts. VNC can provide GUI-based environment for remote administration (including installation -- actually better then such tools as DRAC or ILO) on almost any platform and is very quick and easy to install if you need one ASAP.

Actually monitoring of servers with open source tools is rather easy if the network connectivity is good. There are a several open source tools that are scalable up to a thousand servers without major problems. If we are talking about Unix servers only, then ssh, Perl and Apache server are enough for pretty sophisticated remote monitoring :-).

Software distribution and configuration management are much more complex things. Here enterprise class solution might pay off more quickly. One of the typical configuration management problems that large organizations often have is how to push config files and software updates into multiple boxes after changes in network topology (for example, due to acquisition or divesture). The simplest solution is to have something like a next loop that ssh's into each box and runs a command for us:

for server in `cat machinelist.txt`; do
   echo "running $command on $server"
   ssh user@$server $command; echo " "
done

For more detail see Software Distribution

List of factors that influence number of servers per sysadmin metric

Dr Nikolai Bezroukov


Top Visited
Switchboard
Latest
Past week
Past month

NEWS CONTENTS

Old News ;-)

[Aug 11, 2012] How Many Admins Per User-Computer Have You Seen

December 30, 2009 | Slashdot

TheCarp

Re:Over 9000 (Score:4, Insightful)

I think it comes down to "what do you mean by admins"?

The more you restrict what admins have to support, the more homogenous the environment, the better the ratio. The more you expect from them, the more complex the environment, the more you need.

Another factor to add in is tools. Does every machine have actual remote console? (and I mean console as in, I can sit there and watch the POST console). Do you have build servers and good backups, and tested procedures to do restores if a system needs a total rebuild on the spot?

My group is constantly compared in terms of group size to number of machines. Its maddening since we support 5 different flavors of unix (and VMS), some with ok tools configured and ready, some nearly "hand crafted". Then at least 3 different versions of each of those flavors. We can't seem to get projects approved to fix any of this (and god forbid we did it without a project!)

We are compared against a Windows group, that supports a couple of flavors of Windows, and has had automation tools to schedule and do work remotely setup for years. Of course they can admin more systems with less headcount... they have the tools and environment setup to do it!

Hell it took us almost 2 years to get project approval to set the machines for centralized auth through LDAP... and they wonder why we seem to need so many people for so few machines.

So frankly, I don't think the question has enough information to be answered usefully. There are just too many variables to be able to put up a good estimation of appropriate head count per machine. I can tell you though that standardization, automation, and redundant designs will decrease that head count.

As will properly trained/experienced admins (if we could only get them to send a couple of people to basic sun training we would be way better off... but we can't even get that. We have guys that have been effectively working on the level of entry level admins for years, who have never been able to get management to send them to a class).

-Steve
(who should post anonymously but, on some level hopes they will read this...)

slimjim8094

Depends

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.

RobertM1968

Re:Depends

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.

That last sentence hits the head right on the nail... The numbers really are determined by a lot of factors... if your business revolves around programming and engineering, and thus your workers are from those fields (as opposed to tons of avg computer users in a non computer/technical field), you are less likely to have serious issues that IT needs to address, thus requiring a smaller IT staff.

And of course, what money IT is allowed to spend on initial setup and/or maintenance also determines the staffing size for IT. One can design a system that remote boots from the NIC and reinstalls everything to a machine specific image - or kicks the boot to the HDD if there are no problems making serious non-hardware issues trivial - if the money was there during the initial setup or a big upgrade phase... or one can fix the stuff the old fashioned way and go hands on (which requires more of an IT staff).

Hardware differences also can determine staffing size. One of our customers had a problem with certain AMD XP machines when SP3 came out - required lotsa "hands on" fixing... other of our clients did not have those machines and needed no one and no help. Also, are the machines needed 24/7? Is there mission critical data on them (or no mission critical data anywhere - or mission critical data is on the server)? And so on... inotherwords, there are a ton of factors that determine staffing needs for IT. It could be one person per 10 or one person per 100, etc.

Thus, slimjim8094's statement really does sum it up nicely... Like most of IT, whatever works.

xxxx

Depends :

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^H^H^H^H^H^H the right people.

A good SA can come in and make a lot of these stupid little problems go away, never to return.

These sorts of problems can also be caused by bad management exerting too much control over the admins, or admins with weak people skills trying to please everyone rather than prioritize and do the right thing.

When asked to do something, to you just go ahead and do it? Or do you require things like justifications, business cases, funding, staff, etc? If management can just ask anything of the IT staff, they will do so, and it will feel like you're being walked all over, and that you're overworked. If you have some basic sanity checks and make those requirements before a project can be greenlighted, you'll find that your job can be a lot easier. Doing this also makes planning happen before you get midway through a project and find that different stakeholders have different opinions on what should be done next.

I think you're understaffed.:

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

How many administrators needed per site

comp.unix.admin

Mark Verber

A while ago, I saw a posting calling for input about how many administrators any given site should have. I believe the poster was going to gather the info and publish the stats. I never caught the results. I didn't recall seeing the results either. Well, due to "budget cutbacks", I am in danger of losing my partner/fellow administrator. It seems they think I'm good enough to handle this site all by myself. I'm flattered, but mostly, I'm panicked. I don't think that this is a 1-person job. (I have ~50 Suns on site and ~15 PC's)

So to justify my partner's existance, I need to know: What's the average administrator-to-machine ratio? How about the average administrator-to-USER ratio. Is it different from admin:machine? The administrator to machine ratio that is reasonable depends on what level of service is expected and what responsibilities the sysadm has.

In addition to these two parameters, there are two other issues that will temper the equation:

(1) Single sysadm sites are not very desirable (except financially) in most cases. Is that it is difficult for a single person to have the breadth of knowledge and experience to run a really first class operation, no matter how few machine you have. There will always be some area that a single person will be weak on. It also means that when the sysadm is on vacation, (or gets run over by a bus) the site is vulnerable. Carrying a sky-pager on vacation isn't my idea of fun, and no one can predict when a bus might strike.

(2) The larger a site is, the less people that are needed because you can get a good economy of scale. I have seen at larger sites at a 1:100 sysadm-to-machines where things ran pretty well, although more people would have been nice. I am not sure that there are any hard and fast rules for ratios. The thought of administrating 60 Suns that are more or less alike, a server or three with mainly diskless or dataless clients that are clones of each other wouldn't scare me. That is a pretty manageable task. You would *need* need to be proactive rather than reactive to tasks, make use of existing tools or building some for yourself that permits your work to be multiplied, eg. running something like rdist,target,sup for automatic updates, setting up backups to run automatically (how did we ever live without exebytes), etc, etc. A lot would depend on the level of support your users expect.

The best way to figure out how many people are needed, it to figure out what level of service your users expect and whether the environment helps or hurts the maintainces of software.

Rarely are sysadms doing only UNIX sysadm tasks. Here is a list of things that I find myself doing in addition to the normal "sysadm" tasks like backups, account installs, OS maintaince, etc.

(1) User Services How much hand holding is expected? Some sites have users than are pretty self-sufficient. Other sites have users than need their hands held for everything. Can yours take care of themselves, or do the need/want the admin to do the simplest tasks for them. For example, I have a friend whose users would demand him to do the most basic things for them. Things like moving their files from one directory to another. (Anything that wasn't running the text editor and reading mail was Unix, therefore a job for the adm). This sort of support requires something like a 1:2 ratio. Does the site want you to conduct workshops, prepare extensive local documentation? To what extent is the sysadmin expected to consult on technical issues? Just using Unix, or other realms. For example, lets say your site has heavy users of FrameMaker, TeX, Mathamatica, Common Lisp, C++, X11, PostScript, and Sybase. Is the sysadms suppose to be able to answer detailed questions on all those topics? One person can not be an expert on all these things. Something that many people don't appreciate is that if someone is to consult on a topic area, they need some time to play, experiment, and generally develop in that area.

(2) Diversity of Arch/OS/Setup Of course, each type of machine (arch/OS) that a site has multiples the complexity of the support task, especially if your users expect all machines to behave identically. Each install program will have to be installed N times. OS upgrades will have to be done at least once by hand for each arch/OS, etc, etc. Workstations can be arranged in ways that make it easier, or harder to do distributed administration on. If you can configure all your workstations to use the identical configuration, and have identical tools, etc, it is easy to support a larger number of machines. If each your machines is configured differently: swap sizes different, some diskless, some dataless, some diskful, with different software installed, etc. you are going to have headaches. If you can have one 'prototypical' machine that you can clone from, installs, upgrades, etc can be done pretty easily/automatically. For example, lets say you run a dataless configuration with /, /var, and parts of /usr on a local disk, and everything else accessed via an automounter. Your could have a diskless partition on a server that would boot up, install SunOS on a disk, do your local customizations, and reboot as a newly configured workstation ready for action just by editing your ethers file and a configuration file. If you have to do each machine by hand, you will have to waste a lot of times every time you install a new machine or have to do an OS upgrade.

(3) Software Support? How much PD/freeware software do people want installed, and what level of support are they expecting? Just compiling and installing software doesn't take much time, but often times (and rightly so) a sysadm is expected not only to compile and install new software, but to test the software before people get at it, to debug any problems, port it if necessary, and be a general expert. This takes time which varies with the quality and complexity of the software. Keeping a current version of kermit or perl isn't hard (I wish everyone did as nice a job as Larry has with perl), keeping ontop of g++ takes a quite a bit more time.

(4) Custom Software? Most places not only expect the sysadms to keep the world running, but create tools for the user population when needed. This is understandable, especially in small site where the sysadm might be the decent programmer. If there is this expectation, time must be given for the development process. (5) Site Planning/Admin Overhead How much site planning is the sysadm expected to handle. Are you going to have to know about AC/heating loads and power? How much paperwork is there?

(6) Hardware/Network Maintaince Who crawls through the ceiling to pull wires? How finds the flaky transceiver when the ethernet starts to go crazy? When a terminal or workstation dies do you just call your vendor and wait, or are more creative solutions required. Do you buy all of your peripherals ready to install, or do you save money by purchasing components and do the integration yourself? All of these things take time.

(7) Leading Technology Is the sysadm suppose to anticipate new technology and advise the company as to where they might like to take there environment? Most places I have been, the syadm were expected to have a good feel for what was state of the art and what was looking promising. Not just products, but research. Keeping up with what is going isn't easy. trade rags can give you a picture of what is being sold, but they aren't particularly good at helping people to anticipate what might be in a year or three. Forward looking is often necessary given many sites have a 2-5 year planning and/or depreciation schedule.

Cheers, Mark
More options Jan 16 1991, 10:52 am

Newsgroups: comp.unix.admin
From: ver...@pacific.mps.ohio-state.edu (Mark Verber)
Date: 16 Jan 91 15:52:35 GMT
Local: Wed, Jan 16 1991 10:52 am
Subject: Re: How many administrators needed per site?
Reply to author | Forward | Print | Individual message | Show original | Report this message | Find messages by this author
A while ago, I saw a posting calling for input about how many administrators
any given site should have. I believe the poster was going to gather the
info and publish the stats. I never caught the results.


I didn't recall seeing the results either.


Well, due to "budget cutbacks", I am in danger of losing my
partner/fellow administrator. It seems they think I'm good
enough to handle this site all by myself. I'm flattered, but mostly,
I'm panicked. I don't think that this is a 1-person job. (I have
~50 Suns on site and ~15 PC's)


So to justify my partner's existance, I need to know:
What's the average administrator-to-machine ratio?


How about the average administrator-to-USER ratio. Is it different
from admin:machine?

The administrator to machine ratio that is reasonable depends on what
level of service is expected and what responsibilities the sysadm has.
In addition to these two parameters, there are two other issues that
will temper the equation: (1) Single sysadm sites are not very
desirable (except financially) in most cases. Is that it is difficult
for a single person to have the breadth of knowledge and experience to
run a really first class operation, no matter how few machine you
have. There will always be some area that a single person will be
weak on. It also means that when the sysadm is on vacation, (or gets
run over by a bus) the site is vulnerable. Carrying a sky-pager on
vacation isn't my idea of fun, and no one can predict when a bus might
strike. (2) The larger a site is, the less people that are needed
because you can get a good economy of scale. I have seen at larger
sites at a 1:100 sysadm-to-machines where things ran pretty well,
although more people would have been nice.

I am not sure that there are any hard and fast rules for ratios. The
thought of administrating 60 Suns that are more or less alike, a
server or three with mainly diskless or dataless clients that are
clones of each other wouldn't scare me. That is a pretty manageable
task. You would *need* need to be proactive rather than reactive to
tasks, make use of existing tools or building some for yourself that
permits your work to be multiplied, eg. running something like
rdist,target,sup for automatic updates, setting up backups to run
automatically (how did we ever live without exebytes), etc, etc.

A lot would depend on the level of support your users expect. The
best way to figure out how many people are needed, it to figure out
what level of service your users expect and whether the environment
helps or hurts the maintainces of software. Rarely are sysadms doing
only UNIX sysadm tasks. Here is a list of things that I find myself
doing in addition to the normal "sysadm" tasks like backups, account
installs, OS maintaince, etc.

(1) User Services

How much hand holding is expected? Some sites have users than are
pretty self-sufficient. Other sites have users than need their hands
held for everything. Can yours take care of themselves, or do the
need/want the admin to do the simplest tasks for them. For example, I
have a friend whose users would demand him to do the most basic
things for them. Things like moving their files from one directory to
another. (Anything that wasn't running the text editor and reading
mail was Unix, therefore a job for the adm). This sort of support
requires something like a 1:2 ratio.

Does the site want you to conduct workshops, prepare extensive local
documentation? To what extent is the sysadmin expected to consult on
technical issues? Just using Unix, or other realms. For example,
lets say your site has heavy users of FrameMaker, TeX, Mathamatica,
Common Lisp, C++, X11, PostScript, and Sybase. Is the sysadms suppose
to be able to answer detailed questions on all those topics? One
person can not be an expert on all these things. Something that
many people don't appreciate is that if someone is to consult on
a topic area, they need some time to play, experiment, and generally
develop in that area.

(2) Diversity of Arch/OS/Setup

Of course, each type of machine (arch/OS) that a site has multiples
the complexity of the support task, especially if your users expect
all machines to behave identically. Each install program will have
to be installed N times. OS upgrades will have to be done at least
once by hand for each arch/OS, etc, etc.


Workstations can be arranged in ways that make it easier, or harder to
do distributed administration on. If you can configure all your
workstations to use the identical configuration, and have identical
tools, etc, it is easy to support a larger number of machines. If
each your machines is configured differently: swap sizes different,
some diskless, some dataless, some diskful, with different software
installed, etc. you are going to have headaches. If you can have one
'prototypical' machine that you can clone from, installs, upgrades,
etc can be done pretty easily/automatically. For example, lets say
you run a dataless configuration with /, /var, and parts of /usr on a
local disk, and everything else accessed via an automounter.
Your could have a diskless partition on a server that would boot
up, install SunOS on a disk, do your local customizations, and reboot
as a newly configured workstation ready for action just by editing
your ethers file and a configuration file. If you have to do each
machine by hand, you will have to waste a lot of times every time you
install a new machine or have to do an OS upgrade.

(3) Software Support?

How much PD/freeware software do people want installed, and what level
of support are they expecting? Just compiling and installing software
doesn't take much time, but often times (and rightly so) a sysadm is
expected not only to compile and install new software, but to test the
software before people get at it, to debug any problems, port it if
necessary, and be a general expert. This takes time which varies with
the quality and complexity of the software. Keeping a current version
of kermit or perl isn't hard (I wish everyone did as nice a job as
Larry has with perl), keeping ontop of g++ takes a quite a bit more
time.

(4) Custom Software?

Most places not only expect the sysadms to keep the world running, but create tools for the user population when needed. This is understandable, especially in small site where the sysadm might be the decent programmer. If there is this expectation, time must be given for the development process.


(5) Site Planning/Admin Overhead

How much site planning is the sysadm expected to handle. Are you going to have to know about AC/heating loads and power? How much paperwork is there?


(6) Hardware/Network Maintaince


Who crawls through the ceiling to pull wires? How finds the flaky
transceiver when the ethernet starts to go crazy? When a terminal or
workstation dies do you just call your vendor and wait, or are more
creative solutions required. Do you buy all of your peripherals ready
to install, or do you save money by purchasing components and do the
integration yourself? All of these things take time.


(7) Leading Technology


Is the sysadm suppose to anticipate new technology and advise the company as to where they might like to take there environment? Most places I have been, the syadm were expected to have a good feel for what was state of the art and what was looking promising. Not just products, but research. Keeping up with what is going isn't easy. trade rags can give you a picture of what is being sold, but they aren't particularly good at helping people to anticipate what might be in a year or three. Forward looking is often necessary given many sites have a 2-5 year planning and/or depreciation schedule.


Cheers,
Mark


Newsgroups: comp.unix.admin
From: st...@archone.tamu.edu (Steve Rikli)
Date: 16 Jan 91 20:20:00 GMT
Subject: Re: How many administrators needed per site?
Reply to author | Forward | Print | Individual message | Show original | Report this message | Find messages by this author

In article <1991Jan15.230613.8...@rastro.uucp> bo...@rastro.uucp (christina boyko) writes:

> What's the average administrator-to-machine ratio?


> 1:20, 1:50, 1:100? (This assumes that the users do absolutely no administrative tasks.---They don't even have root access.) (My personal feeling is 1:25 at most, but management thinks 1:60....)

> How about the average administrator-to-USER ratio. Is it different from admin:machine?

Here in the College of Architecture's Visualization Lab, we have

- a sun4 server
- ~15 sun3's
- ~10 sparc's
- 2 NeXT's
- 8 SGI's of various flavors
- 40-odd PC's
- 30-odd Mac's


There are approximately 100 users with active accounts, but I would
say only about 20-30 are regular users.


Of the Mac's and PC's, only a handful have direct connections to the
server, so they are not directly our concern. In addition, there
are various printers, scanners, and other peripherals that fall under
our jurisdiction.


The "we" above consists of a full-time system administrator, and a
half-time assistant system-administrator (who is a grad student).
There is also a full-time staff member to handle most Mac and PC
problems, so there will naturally be some overlap.


_________________________________________________________
/ st...@archone.tamu.edu / Work: (409) 845-3465 /
/ srr2...@sigma.tamu.edu / Home: (409) 696-0910 /
/___________________________/____________________________/
/ Steve Rikli, Assistant System Manager /
/ Visualization Lab, College of Architecture /
/ Texas A&M University /
/ College Station, TX 77844 /
/________________________________________________________/


Ed Anselmo View profile
More options Jan 16 1991, 6:00 pm

Newsgroups: comp.unix.admin
From: anselmo...@CS.YALE.EDU (Ed Anselmo)
Date: 16 Jan 91 23:00:27 GMT
Local: Wed, Jan 16 1991 6:00 pm
Subject: Re: How many administrators needed per site?

Yale Computer Science has a Facility composed of 1 Facility Director, 1 Manager of Development, 2 Sr. Systems Prog. (I'm one of them), 1 Manager of Operations, 1 Operations programmer, 1 Operations staff.

This is down from 4 Sr. Programmers, 2 Operations programmers, and 2 Operations staff 2 years ago. And they've cut out the weekend backup operators too.

We manage about 150 Suns + miscellaneous other machines (Connection Machine, Intel Hypercube, Encore Multimax, Sequent Symmetry, some IBM-RT's, IBM RS/6000, DEC-5000's). Maybe 200 machines in all.

The user community is maybe 600 (undergrads, grads, faculty, and staff).

I was the sole support person in my last job (5 Suns, 40 IBM-PCs, 10 Macs). They expected me to do development when I could barely keep all the machines and printers alive from day to day. Yeah, sure.

>> (1) User Services

Users get daily backups, except on weekends. A staff person is on call on weekends for extreme emergencies, otherwise we're a 9-to-5 weekdays operation.

We seem to be fair game for questions on anything that falls under the "supported" category: TeX, X, mail, news, networking, etc. Each the development staff is at least dimmly aware of every major software package we support, though "The Other Guy" handles TeX, I field news, mail, and networking questions, and we split up the X support.

>> (2) Diversity of Arch/OS/Setup

Since CS switched to all Sun Sparcstations, maintenance has become much easier. 2 years ago, the facility supported 4 workstation architectures: Apollo, HP, IBM, and Sun. Now it's pretty much just Sun, though we just got in 10 DECstations, and a similar number of Macs.

>> (3) Software Support?

The raging issue amongst the Powers That Be is just what software is to be supported and at what levels. At our current staffing level, it's basically impossible to do feature enhancement; fixing serious bugs is doable; but mostly, we just install/port the software, and rely on others to provide us with fixes.


>> (4) Custom Software?

We support several locally written pieces of software, like the
magical software that hides all the userids in CS behind the
"lastname-firstname" alias for the purposes of news and mail, the User
Database program (manages accounts/uids/mailing-lists), the "autodump"
program that manages the file backups, an editor, and a mail reader.


>> (5) Site Planning/Admin Overhead


The facility director gets stuck with this job.


>> (6) Hardware/Network Maintaince


We have on-site Sun hardware support. Plus a relatively new building
with professionally installed thicknet. I did my time in the Midnight
Wiring Crew at my last job. Ugh.

We let our hardware technician take care of most things. (Thankfully,)
he yells at me when I start poking around the multiports in the comm.
closets. So I let him do the work.

>> (7) Leading Technology


"I'm reading news so that I can keep up with new technologies. YEAH,
that's it, NEW TECHNOLOGIES, that's the ticket."
--
Ed Anselmo anselmo...@cs.yale.edu {harvard,cmcl2}!yale!anselmo-ed

Reply to author Forward

You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.

F. L. Charles Seeger III View profile
More options Jan 17 1991, 1:57 am

Newsgroups: comp.unix.admin
From: see...@thedon.cis.ufl.edu (F. L. Charles Seeger III)
Date: 17 Jan 91 06:57:57 GMT
Local: Thurs, Jan 17 1991 1:57 am
Subject: Re: How many administrators needed per site?
Reply to author | Forward | Print | Individual message | Show original | Report this message | Find messages by this author
In article <1991Jan15.230613.8...@rastro.uucp> bo...@rastro.uucp (christina boyko) writes:


| A while ago, I saw a posting calling for input about how many administrators
| any given site should have. I believe the poster was going to gather the
| info and publish the stats. I never caught the results.

I, too, would like to see more information on this subject, especially from
big name CS departments. Frankly, I'm looking for ammunition for improving
the situation here.


| Well, due to "budget cutbacks", I am in danger of losing my
| partner/fellow administrator. It seems they think I'm good
| enough to handle this site all by myself. I'm flattered, but mostly,
| I'm panicked. I don't think that this is a 1-person job. (I have
| ~50 Suns on site and ~15 PC's)


Depends on more details, but it is very unlikely to be a 1-person job.


| So to justify my partner's existance, I need to know:
| What's the average administrator-to-machine ratio?


Here is brief description of the situation here. I consider this to be
a very bad (understaffed) situation. I don't know if this will help
bolster your position or not. I don't think your employers will want
their system staff to be as overloaded as we are. Everyone suffers.


We have one Senior Sys. Programmer (me), one Sys. Programmer, and one
Electronic Technician. We have about 10 part time student employees
(20 hrs/wk for an assistant for the E-Tech, 20 hrs/wk for backups, about
25 hrs/wk for software and documentation support, and about 105 hrs/wk for
manning our workstation lab). We have no secretarial or other administrative
support assigned for managing our facilities. The second Sys. Prog.
position has existed only since the first of last July.


The E-Tech spends much of his time handling administrivia such as shipping
and receiving, inventory control, purchasing supplies, handling support
contracts. I do all the vendor interaction for purchasing new equipment.
Accounting and other business functions in our department office are a
sick joke.


We have 80+ Suns, 7 HPs, 4 RTs, 1 RS6k, 1 Ardent, 1 Gould PowerNode, various
terminals, several PCs and two Macs. We have 27 faculty members, a passwd
file with about 1000 entries, and about 6 GB of user disk space with 3 more
on the way. We also have some non-standard equipment, such as two different
Transputer systems and two different specialized computer vision systems.
Our /local partition consumes much more space than all of SunOS (including
OW2 and the unbundled compilers).


We also run Florida's biggest news feed and uucp machine. Mail sent to
postmas...@ufl.edu goes into my mailbox. Thank dmr that someone else
handles assigning IP addresses on this campus.


Enough griping about the situation here. How do *real* CS departments
stack up? We've heard from UMd Eng (but not CS), Yale, Ohio State Physics
(but not CS). What about Stanford, UCB (or UC*), Michigan, Illinois, CMU,
MIT, Harvard, Purdue, PSU, Utah, GaTech, Texas, etc.?


Bombard me with useful ammunition, please.


Regards,
Chuck
--
Charles Seeger E301 CSE Building Office: +1 904 392 1508
CIS Department University of Florida Fax: +1 904 392 1220
see...@ufl.edu Gainesville, FL 32611-2024


Reply to author Forward

You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.

Steve Romig View profile
More options Jan 17 1991, 3:20 pm

Newsgroups: comp.unix.admin
From: ro...@brachiosaur.cis.ohio-state.edu (Steve Romig)
Date: 17 Jan 91 20:20:36 GMT
Local: Thurs, Jan 17 1991 3:20 pm
Subject: Re: How many administrators needed per site?
Reply to author | Forward | Print | Individual message | Show original | Report this message | Find messages by this author

> Enough griping about the situation here. How do *real* CS departments
> stack up? We've heard from UMd Eng (but not CS), Yale, Ohio State Physics
> (but not CS).


We've got ~230 diskless SLCs, served by 21 Sun 3/180 file servers (and
a 4/280 and a 4/330); ~10 diskless HP somethings served by 1 HP file
server; 4 Pyramids; 1 Multimax; 1 Butterfly; 300+ Macs; and some odds
and ends. There are roughly 9 different hardware/software platforms
that we currently support (sun3 running sunos 4.1, sun4 running sunos
4.1, etc). This is all connected through 1 main ethernet (our
backbone), 24+ Ethernet subnets and a bunch of Appletalk stuff that I
don't want to know anything about.

Our users: roughly 1700-1800. 45 faculty, 200 grad students, rest are
undergrads or guest acounts. These facilities are for instruction and
research in the Computer and Info Sciences Department at OSU. That
count doesn't include the students using the Macs in the low level
courses, which is probably another 1500 folks or so.


Our staff: We're split into 3 parts: software, hardware and
operations.


Software staff deals with software development and systems
installation, maintenance and bug tracking/fixing. Consists of 8
full time folks (6 Unix, 1 Mac, 1 Unix/parallel research support)
and 8 part time folks (grads and undergrads). We're all (but 1)
general Unix folks, though we each tend to specialize in different
areas (X, postscript, printers and text processing stuff,
networks, mail, news, strange languages, ntp, nameservers, etc).


Hardware staff deals with hardware install, maintenance, advice on
upgrades and etc. We do almost all of our Sun (and I think most
of our Mac support) in house, at the board component level. We
also do most of our own peripheral integration (select, buy and
install disks, tapes, etc). The rest is through support contracts
with the vendors. Hardware consists of 2 full time folks and
something like 6 part time folk. Oh, they take care of the nets
too.


Operations staff deals with keeping things running: acount
installation, maintenance, file system stuff (creating and
maintaining user and project directories, backups, restores,
handling common problems and emergencies) annnnd they are
stationed in the labs when they are open to handle problems,
answer questions, and keep people from walking away with or
destroying machines. 1 full time person, 35 part time.


We don't do much in the way of course-ware development, but do do alot
of consulting type things with our various users. The software staff
(especially) is expected to and is trying to do more development type
work (make new/better sysadmin tools, better user interface type
things, etc), though our main "purpose" is to keep things running and
reasonably up to date.


Lessons we've learned:

Keep everything as much the same as possible. All of our Sun
clients are clones of a master copy, all of the servers are clones
of a master server, all of the Pyramids look alike, etc.


Reduce the number of platforms as much as possible. We used to
have something like 13 platforms, we're down to 9, and may soon be
down to 7 if we lose the Pyramids...In my mind, though reducing
the number of platforms is nice, you have to balance that against
having a rich environment, which is also nice.


Localize local changes to /usr/local (or some scheme like that) as
much as possible, which makes upgrades easier. Try to refrain
from hacking on and reinstalling local versions of things in /bin,
/usr/ucb, and so on.


Diskless workstations are your friend, as long as you have enough
memory on them. It takes me about 2 hours to install a new copy
of / on all 220+ diskless SLCs, including shutting them down,
copying the stuff and bringing everything up again.


You have to strike a balance between keeping things up to date and
spending too much time keeping things up to date. I try to settle
on a SunOS release that seems reasonably stable and stay there for
a long time, for example. We were at SunOS 3.5.1 for a very long
time. We're at 4.1 now, and I'm still searching for a point of
stability...:-)


Build tools to do things, rather than doing it "by hand" - if you
do something once, you'll do it again.


Beg, borrow and steal (only kidding) software from others when you
can.


(Mark came up with the

Number of Admin's per Server - Database Forum

02-05-2005 | Database Forum

Re: Number of Admin's per Server
V Crenshaw writes:

> Have you seen in any of the industry mags with a survery or article that
> talks about the
> optimum number of unix or AIX admins per server?

Unless the workload is very heavy, a single administrator should be
optimal. For mission-critical machines, two people with root access
would probably be wise (although only one might normally carry out all
the administration duties).

For large systems and in cases where responsibility must be divided for
security reasons, you might want several administrators with different
duties. The UNIX architecture doesn't lend itself to this, though,
since you're either root or you're not, so it's all or nothing as
privileges go. There are some creative workarounds, but they involve
allowing ordinary users to run a few privileged programs; the only real
administrators are those with the root password, and anyone with the
root password can do anything.

One of the problems I've always had with UNIX is this all-or-nothing
sysadmin privilege, but it's a problem shared by a great many operating
systems.

--
Transpose hotmail and mxsmanic in my e-mail address to reach me directly.


#3
02-05-2005, 07:57 PM
Database Administrator
Database Bot

Join Date: Sep 2009
Posts: 1,236,254

Re: Number of Admin's per Server
I guess I wasn't clear in what I was asking. We have thousands of midrange
unix
servers: SUN, AIX, PYRAMID and HP in the company I work for.

I'm asking how many AIX servers should one admin be responsible for - max.

I'll take opinions, but I'm really asking if anyone has seen any written
opinions from industry
writers and magazines.

"Mxsmanic" wrote in message
news:epga01511pms0jve0vdislg6809m2v51ck@4ax.com...
>V Crenshaw writes:
>
>> Have you seen in any of the industry mags with a survery or article that
>> talks about the
>> optimum number of unix or AIX admins per server?
>
> Unless the workload is very heavy, a single administrator should be
> optimal. For mission-critical machines, two people with root access
> would probably be wise (although only one might normally carry out all
> the administration duties).
>
> For large systems and in cases where responsibility must be divided for
> security reasons, you might want several administrators with different
> duties. The UNIX architecture doesn't lend itself to this, though,
> since you're either root or you're not, so it's all or nothing as
> privileges go. There are some creative workarounds, but they involve
> allowing ordinary users to run a few privileged programs; the only real
> administrators are those with the root password, and anyone with the
> root password can do anything.
>
> One of the problems I've always had with UNIX is this all-or-nothing
> sysadmin privilege, but it's a problem shared by a great many operating
> systems.
>
> --
> Transpose hotmail and mxsmanic in my e-mail address to reach me directly.


#4
02-06-2005, 06:00 AM
Database Administrator
Database Bot

Join Date: Sep 2009
Posts: 1,236,254

Re: Number of Admin's per Server
In article ,
"V Crenshaw" wrote:

> I guess I wasn't clear in what I was asking. We have thousands of midrange
> unix servers: SUN, AIX, PYRAMID and HP in the company I work for.
>
> I'm asking how many AIX servers should one admin be responsible for - max.
>
> I'll take opinions, but I'm really asking if anyone has seen any written
> opinions from industry writers and magazines.

"That depends" (the standard answer to overly general and not well
understood by the the Original Poster)

There's no hard and fast formula and most PHBs are looking for one
thinking sysadmins are like widgets. Guess what, they aren't. Good
ones know the systems they run and keep them running well, like a Kabuki
stagehand who's seen on stage but not really seen.

Factors that affect sysadmin ratio:

- environment and service level expectations (7x24 production under
tight change controls with measured up-time (4 Nines 5 or better=130s
downtime/month), development, or available during standard business
hours)
- workload (heavily used mission-critical with HA, critical -
infrastructure system like directory services or web, or developer
systems),
- utilization patterns (system load creeps toward double digits just
before lunch, dips, then goes back up until about 5pm),
- applications care and maintainance ("Oracle is running slow", "Why is
the company web page so slow today", "I need files restored that were on
my PC")
- hardware configuration (disk space requirements, memory utilization,
older or questionable hardware)
- software configuration (OS version and patch levels)
- degree of infrastructure automation and monitoring (if there isn't
any, you're in deep dodo!)

If you have groups of systems that are similar running about the same
hardware, then one or two people could handle the similar systems (say
the DB2 systems). The more different the systems are, the less overlap
you'll be able to depend on. It also depends on the degree of
cross-pollination that occurs between the various sysadmins. If they
are a cohesive group and work together as a team, then you get synergy.
If they're scatter all over the enterprise, there's probably very little
overlap, except for various infrastructure services like the Datacenter
and Network people that can mandate various things.

The Practice of System and Network Adminstration by Thomas Limoncelli
and Christine Hogan (ISBN 0201702711) has an excellent discussion of
this whole topic from a general, system-independent architecture.

--
DeeDee, don't press that button! DeeDee! NO! Dee...

[SAGE] Servers per SysAdmin by Barnette, Steve L

8 Oct 2001

I am in a staffing discussion on number of SysAdmins needed for our servers.
The suits are quoting some Gardner Group article stating one SysAdmin per
hundred servers. That sounds high to me, I was wondering what all of you
think?


Steve Barnette
Sr. Systems Engineer
-------------------------------------------------------
# perl -e 'print $i=pack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);

Re [SAGE] Servers per SysAdmin

As we found out at OSDN, 2 sysadmins can handle 200 servers but you'll run
them ragged doing so, I would say, honestly 50 per sysadmin, 30 if theres
an on-call schedule.

I know, I ended up in the hospital from the stress.

-Trish

RE [SAGE] Servers per SysAdmin

how many sysadmins does it take to screw in a light bulb.... :)

this is very misleading. it depends what the servers are doing and the
experience of the administrator(s). 100 homogeneous servers all running the
same operating system, with centralized authentication and software
distribution mechanisms is very manageable with one or two senior
administrators. 20 servers running different operating systems on different
hardware platforms with varying capacity and performance requirements can be
more difficult to manage, and may require two or three administrators. at
my current job, i alone administer 30 servers. a couple years ago i was at
the university of michigan, where a team of myself and 5 other
administrators managed the same number of machines. the complexity and
dynamic nature of the services we were offering (login services to 80,000
users!) required us to have more staff.

the moral of the story is, upper management cannot base their staffing
requirements on a simple formula.

-jeff

Re [SAGE] Servers per SysAdmin

Barnette, Steve L <steve_b3@corp.earthlink.net> writes:

> I am in a staffing discussion on number of SysAdmins needed for our
> servers. The suits are quoting some Gardner Group article stating one
> SysAdmin per hundred servers. That sounds high to me, I was wondering
> what all of you think?

I think it depends highly on how homogenous your servers are. If you only
have one to five servers for each application, a hundred servers is
probably way too high; if the hundred servers are a single compute cluster
composed of completely identical machines, with the appropriate automation
a single sysadmin could easily handle that and more.
--
Russ Allbery (rra@stanford.edu) <http://www.eyrie.org/~eagle/>

Number of Admin's per Server - Database Forum

V Crenshaw writes:

> Have you seen in any of the industry mags with a survery or article that
> talks about the
> optimum number of unix or AIX admins per server?

Unless the workload is very heavy, a single administrator should be
optimal. For mission-critical machines, two people with root access
would probably be wise (although only one might normally carry out all
the administration duties).

For large systems and in cases where responsibility must be divided for
security reasons, you might want several administrators with different
duties. The UNIX architecture doesn't lend itself to this, though,
since you're either root or you're not, so it's all or nothing as
privileges go. There are some creative workarounds, but they involve
allowing ordinary users to run a few privileged programs; the only real
administrators are those with the root password, and anyone with the
root password can do anything.

One of the problems I've always had with UNIX is this all-or-nothing
sysadmin privilege, but it's a problem shared by a great many operating
systems.

--
Transpose hotmail and mxsmanic in my e-mail address to reach me directly.

Number of Admin's per Server - Database Forum

In article ,
"V Crenshaw" wrote:

> I guess I wasn't clear in what I was asking. We have thousands of midrange
> unix servers: SUN, AIX, PYRAMID and HP in the company I work for.
>
> I'm asking how many AIX servers should one admin be responsible for - max.
>
> I'll take opinions, but I'm really asking if anyone has seen any written
> opinions from industry writers and magazines.

"That depends" (the standard answer to overly general and not well
understood by the the Original Poster)

There's no hard and fast formula and most PHBs are looking for one
thinking sysadmins are like widgets. Guess what, they aren't. Good
ones know the systems they run and keep them running well, like a Kabuki
stagehand who's seen on stage but not really seen.

Factors that affect sysadmin ratio:

- environment and service level expectations (7x24 production under
tight change controls with measured up-time (4 Nines 5 or better=130s
downtime/month), development, or available during standard business
hours)
- workload (heavily used mission-critical with HA, critical -
infrastructure system like directory services or web, or developer
systems),
- utilization patterns (system load creeps toward double digits just
before lunch, dips, then goes back up until about 5pm),
- applications care and maintainance ("Oracle is running slow", "Why is
the company web page so slow today", "I need files restored that were on
my PC")
- hardware configuration (disk space requirements, memory utilization,
older or questionable hardware)
- software configuration (OS version and patch levels)
- degree of infrastructure automation and monitoring (if there isn't
any, you're in deep dodo!)

If you have groups of systems that are similar running about the same
hardware, then one or two people could handle the similar systems (say
the DB2 systems). The more different the systems are, the less overlap
you'll be able to depend on. It also depends on the degree of
cross-pollination that occurs between the various sysadmins. If they
are a cohesive group and work together as a team, then you get synergy.
If they're scatter all over the enterprise, there's probably very little
overlap, except for various infrastructure services like the Datacenter
and Network people that can mandate various things.

The Practice of System and Network Adminstration by Thomas Limoncelli
and Christine Hogan (ISBN 0201702711) has an excellent discussion of
this whole topic from a general, system-independent architecture.

--
DeeDee, don't press that button! DeeDee! NO! Dee...

==

Number of Admin's per Server - Database Forum

In comp.unix.admin Michael Vilain :
> In article ,
> "V Crenshaw" wrote:


>> I guess I wasn't clear in what I was asking. We have thousands of midrange
>> unix servers: SUN, AIX, PYRAMID and HP in the company I work for.
>>
>> I'm asking how many AIX servers should one admin be responsible for - max.
>>
>> I'll take opinions, but I'm really asking if anyone has seen any written
>> opinions from industry writers and magazines.


> "That depends" (the standard answer to overly general and not well
> understood by the the Original Poster)

Or just refused, cowardly unwilling to except reality.

> There's no hard and fast formula and most PHBs are looking for one
> thinking sysadmins are like widgets. Guess what, they aren't. Good
> ones know the systems they run and keep them running well, like a Kabuki
> stagehand who's seen on stage but not really seen.

100% ack. There's no way of determining this in a general way,
even if management would like to. It could be 150 or more systems
in an easy environment or just 25, with a bunch of developers on
them, keeping the admin pretty busy.

[..]

I found some good information in
. Granted it
assumes the org has bought into using Tivoli, but there's some good
content in there that can give rise to thoughts that should be applied
to an organization, irrespective of whatever tools (or tool suites) are
actually deployed. They provide some good analysis of the different
trade spaces that occur in any environment.

Recently, I was able to play around with AIX's System Manager. I'm not
saying it's any kind of silver bullet, but IMO they've done a nice job
of allowing multiple nodes to be centrally managed. I haven't played
with it extensively, but I'd like to see a similar interface work in a
heterogenous environment, for at least the portions which are common
across many a Unix platform.

Regards,
Jon

How many administrators needed per site - comp.unix.admin Google Groups

This is a discussion on Number of Admin's per Server within the unix-admin forums in Operating Systems category; I'm new to the group. I've looked at the FAQ's and archives and can't find anything referencing this so I have a question. Have you seen in any of the industry mags with a survery or article that talks about the optimum number of unix or AIX admins per server? I know servers have different levels of complexity etc. but I'm looking for some data to see what is generally done in the real world. If you recall seeing something and are not really sure, would you tell me where you likely saw it, ie. ...

I'm new to the group. I've looked at the FAQ's and archives and can't find anything referencing this so I have a question.

Have you seen in any of the industry mags with a survery or article that talks about the optimum number of unix or AIX admins per server?

I know servers have different levels of complexity etc. but I'm looking for some data to see what is generally done in the "real world."

If you recall seeing something and are not really sure, would you tell me where you likely saw it, ie. what
mags you might have seen it in or website etc.?

VCrenshaw

02-05-2005, 06:09 PM

Re: Number of Admin's per Server

V Crenshaw writes:

> Have you seen in any of the industry mags with a survery or article that
> talks about the
> optimum number of unix or AIX admins per server?

Unless the workload is very heavy, a single administrator should be
optimal. For mission-critical machines, two people with root access
would probably be wise (although only one might normally carry out all
the administration duties).

For large systems and in cases where responsibility must be divided for
security reasons, you might want several administrators with different
duties. The UNIX architecture doesn't lend itself to this, though,
since you're either root or you're not, so it's all or nothing as
privileges go. There are some creative workarounds, but they involve
allowing ordinary users to run a few privileged programs; the only real
administrators are those with the root password, and anyone with the
root password can do anything.

One of the problems I've always had with UNIX is this all-or-nothing
sysadmin privilege, but it's a problem shared by a great many operating
systems.

--
Transpose hotmail and mxsmanic in my e-mail address to reach me directly.

02-05-2005, 07:57 PM

Re: Number of Admin's per Server
I guess I wasn't clear in what I was asking. We have thousands of midrange
unix
servers: SUN, AIX, PYRAMID and HP in the company I work for.

I'm asking how many AIX servers should one admin be responsible for - max.

I'll take opinions, but I'm really asking if anyone has seen any written
opinions from industry
writers and magazines.

"Mxsmanic" wrote in message
news:epga01511pms0jve0vdislg6809m2v51ck-at-4ax.com...
>V Crenshaw writes:
>
>> Have you seen in any of the industry mags with a survery or article that
>> talks about the
>> optimum number of unix or AIX admins per server?

>
> Unless the workload is very heavy, a single administrator should be
> optimal. For mission-critical machines, two people with root access
> would probably be wise (although only one might normally carry out all
> the administration duties).
>
> For large systems and in cases where responsibility must be divided for
> security reasons, you might want several administrators with different
> duties. The UNIX architecture doesn't lend itself to this, though,
> since you're either root or you're not, so it's all or nothing as
> privileges go. There are some creative workarounds, but they involve
> allowing ordinary users to run a few privileged programs; the only real
> administrators are those with the root password, and anyone with the
> root password can do anything.
>
> One of the problems I've always had with UNIX is this all-or-nothing
> sysadmin privilege, but it's a problem shared by a great many operating
> systems.
>
> --
> Transpose hotmail and mxsmanic in my e-mail address to reach me directly.


02-06-2005, 06:00 AM

Re: Number of Admin's per Server
In article ,
"V Crenshaw" wrote:

> I guess I wasn't clear in what I was asking. We have thousands of midrange
> unix servers: SUN, AIX, PYRAMID and HP in the company I work for.
>
> I'm asking how many AIX servers should one admin be responsible for - max.
>
> I'll take opinions, but I'm really asking if anyone has seen any written
> opinions from industry writers and magazines.

"That depends" (the standard answer to overly general and not well
understood by the the Original Poster)

There's no hard and fast formula and most PHBs are looking for one
thinking sysadmins are like widgets. Guess what, they aren't. Good
ones know the systems they run and keep them running well, like a Kabuki
stagehand who's seen on stage but not really seen.

Factors that affect sysadmin ratio:

- environment and service level expectations (7x24 production under
tight change controls with measured up-time (4 Nines 5 or better=130s
downtime/month), development, or available during standard business
hours)
- workload (heavily used mission-critical with HA, critical -
infrastructure system like directory services or web, or developer
systems),
- utilization patterns (system load creeps toward double digits just
before lunch, dips, then goes back up until about 5pm),
- applications care and maintainance ("Oracle is running slow", "Why is
the company web page so slow today", "I need files restored that were on
my PC")
- hardware configuration (disk space requirements, memory utilization,
older or questionable hardware)
- software configuration (OS version and patch levels)
- degree of infrastructure automation and monitoring (if there isn't
any, you're in deep dodo!)

If you have groups of systems that are similar running about the same
hardware, then one or two people could handle the similar systems (say
the DB2 systems). The more different the systems are, the less overlap
you'll be able to depend on. It also depends on the degree of
cross-pollination that occurs between the various sysadmins. If they
are a cohesive group and work together as a team, then you get synergy.
If they're scatter all over the enterprise, there's probably very little
overlap, except for various infrastructure services like the Datacenter
and Network people that can mandate various things.

The Practice of System and Network Adminstration by Thomas Limoncelli
and Christine Hogan (ISBN 0201702711) has an excellent discussion of
this whole topic from a general, system-independent architecture.

--
DeeDee, don't press that button! DeeDee! NO! Dee...

Number of servers per server administrator..

LinkedIn Answers

Good Answers (5)

Mark Verber, Director of IT at Stanford University

see all my answers

Best Answers in: Staffing and Recruiting (1)... see more, Computers and Software (1) see less

This was selected as Best Answer

There is no magic ratio. There are a large number of factors which effect this. I wrote a short article about this something like 17 years ago. I slightly update version of this article is linked. One of these days I might get around to updating it. It won't give you hard answers, but it will help you ask better questions.

--mark

Links:

Samdani Basha, Vice President at GE

see all my answers

Hi Prasad,
I am not aware of any Industry Benchmarks / Best Practices in this area but I would approach this by analysing / considering the below mentioned points

1. Historical Ananlysis of Incident Count ( Incidents, Problems, Change Request )
2. Appox Effort Estimation ( Man Mins ) by Incident
3. Concurrency of Incidents
4. Service Levels that you signed up for
5. Productivity Committments that the your Customer Expects over the life cycle of the Contract
6. Operational Excellence Road Map that your Process Improvement Team comes up with ( Stuff like MTBF, Value Add - Non Value Add Analysis etc )
7. Standard Staffing Rules ( Hours per shift, Leave Policy, Holiday Calendar, Back Up Requirememts, On Call Requirements )
8. Pricing Model that your Organization follows ( Incident Based, Fixed Price, SLA Based, FTE Based )

Adnan Rafik [adnan@itselekt.com], CITP [MBCS], Top of The Mark Volunteer Award Winner, IT Pro Community Leader, Prof. Speaker, Technical Event Organizer

see all my answers

Mark, Dave, Louis, Richard all have given the answers and yes there is not a rule of thumb.

To me why do have this question in your mind? what made you think like this. Look at the other side of the picture... what do you want to achieve from this ... do you want to reduce the cost to pay to the server managers/administrators?

Is uptime more important for you?
Does your server system admin dependent?
What applications are you running and how complex and critical are they?

have these questions yourself and find the answers within you.

David Kramer, Owner, Cooperative Computing

see all my answers

Prasad,

It appears you have asked a couple of different questions. Did you want to know how many Technical Staff to man a 24x7x365 Server Administration Function ??? If so then if you have a standard 2 week time off and 12 Standard Holidays and you have 12 hour shifts then use 8 Bodies per position you wish to staff. You can determine the number of positions by common work load factors. The number of "incedents" per device depends on what you have the "SA" doing. Feel free to contact me if yuo want to develop a more complex model and associated costs averaged by application types (Applicatoin refers to OS only, Oracle Database, Exchange, etc) - My email is david.kramer@cooperativecomputing.net

Some standard "Functional" assement methods can be seen at the attached link. I have more detailed methods and practical models for use in commercial applications where "Mission Critical" has a different meaning.

"HT"

Links:

Phillip Maddocks, Information Technology and Professionl Services Consultant

see all my answers

Best Answers in: Currency Markets (1)... see more, Computers and Software (1), Information Security (1) see less

Prasad,

All the previous answers are correct, there is no "standard" in my experience as well as looking at OS you also need to be looking at the applications in use and what is necessary to maintain and support those applications, you don't specify platform, which again is important as many tasks could be automated, so again the level of automation may affect numbers required, you may need specialised admin staff or may be able to have staff "on-call" this may depend on the experience of the staff concerned. If you are attempting to introduce 24x7 support then I would suggest that you document the current sys admin tasks along with timings, frequency, numbers of users etc etc in the particular timeframe and then seeing how the staff numbers balance out i.e. if you need 4 staff during the "busiest timeframe" can you manage with 2 on quieter periods etc. if business expands then you raise the business case to increase the number of staff.
Hope this helps

Nagesh B R

Managed Support functions in an Enterprise environment. Versatile: small to large multifunctional and multiple teams.

see all my answers

By now you know that there is no rule of thumb for the number of System Administrators (SA). The answers are here and can be found by you. Here is my 2p worth suggestion:
1. Case where the company has an existing and similar infrastructure or Data Centre(DC)
* Get all the info suggested in the different answers here and you can arrive at your own ball-park figure (ask about the different platforms they aare supporting and the different applications they are supporting, what is the inflow of issues/requests or tickets and how do they handle it); it would give you enough ideas to start with. Maybe you could then run your estimates with the existing stakeholders of the DC.
* Get the lead administrator/supervisor of your distributed infrastructure to talk/visit them and build a rapport that could be useful at a later time to help solve a more complex issue...
2. If the infrastructure you are planning to set up is new:
* Try to standardize on platforms as much as possible (question your limits): the more diverse, the more are the issues.
* The vendors are eager to sell: ask them suggestions in administering the systems: ask them if they would be willing to help train your SA(s) at a subsidised (or free) cost. Go for a good combination of warranty-support that could reduce the burden of SA on your org. (sometimes vendors even provide on-site support depending on the quantum of business).
3. The 24x7 administration: determine what is the "lean" time and what is "peak" time. The ration of requirement at lean time, therefore may be a fraction: you may require 30-40% of the number of SA. Have a contingency plan for the Lean period: maybe some SA could be on-call with a slew of additional benefits if called....This way you may employ only 1.7X the peak SA for all three shifts: in place of 3X SA for 3-shifts. You might even play with shifts later: define L1/l2 support etc. and an escalation mechanism with SLAs etc. and a general or peak shift..as you grow the organization.

Hope that it helps.

Clarification added 8 months ago:

One more thing: a good ticketing or incident management system with a shared or web based detailed documentation-help-FAQ really helps the SA team and even the end users in the case of a high volume and/or 24x7 environment! There are many free or reasonably priced ticketing systems available today that are almost as good as the more expensive (hundreds or even thousands of times more expensive) ones from large brand-names.

Daver Mahiar, Head Innovation Management (AP) at Bayer Business Services

see all my answers

If there is a remote mangement tool to access the same. then the number is around 200-300.
if no software is present and physical present is required, then not more than 50-100 servers per server administrator.

also i am assuming these servers to be windows servers.
in case of Unix boxes, more servers can be managed even without management software, as patches to be applied are less often and they hardly need to be rebooted. so zero admin scenarios can exisit with unix boxes.

hope this answers. for more information on such services, plase get in touch with me. we provide infrastructure services to Construction companies even in remote areas like LEH.

Louis Rosas-Guyon

louis(at)r2computing.com; Business Technology Expert, Blogger, Philanthropist & Dad

see all my answers

Best Answers in: Government Policy (3)... see more, Web Development (3), Mentoring (2), Internationalization and Localization (2), Advertising (2), Project Management (2), Supply Chain Management (2), Computers and Software (2), Software Development (2), Using LinkedIn (2), Regulation and Compliance (1), Treaties, Agreements and Organizations (1), Contracts (1), Internet Marketing (1), Graphic Design (1), Business Analytics (1), Corporate Governance (1), Change Management (1), Organizational Development (1), Planning (1), Commodity Markets (1), Personal Investing (1), Career Management (1), Business Plans (1), Starting Up (1), Blogging (1), E-Commerce (1), Enterprise Software (1) see less

I hate to break it to you but there is no hard and fast formula for this one. It all boils down to complexity of the environment and the number of users that must be administered. In my time I have seen networks with 1500 users and 75 servers run by a single administrator but I have also seen networks with 5 servers and 50 users run by three admins. It all depends on the duties of the admin and the environment. Sorry I couldn't be more help, but trying to develop a formula for this would lead to IT suicide.

posted 8 months ago | Flag answer as...

Richard Rothwell

Owner, M6-IT CIC

see all my answers

I have to agree that no rule of thumb does - or can - exist. What the organisation should be doing is looking at its systems, analysing its requirements and coming up with a strategy to reduce the demand that this places on the technical staff.

Of course, they have to remember that they cannot ask the techies how much work is involved, neither can they ask their non-technical managers. One will not know, the other will be unreliable in their answer. Please decide which is which for yourself, there are hints in the link below. Only an external expert can collect the data and give an accurate answer. The brief for the expert should include developing a strategy to reduce the demand on technical support.

--
Richard

Links:

Deepak Anand, Wipro Technologies

see all my answers

As everyone said - "No thumb rule here", I too agree to it. With my past experience, both in a Data Center and handling remotely (RIM), I can tell you these infrastructures (both Compute and Networking devices) do take lot of attention. Many a times in Unix boxes you don't need much people to take care of them, but in Windows you need many of them. And if in case it has got multiple applications running on it, then you need to have lot of Sys Admin. To make your life easy, you will be on safe side to assign 40-75 box per Admin, but do ensure that Application Support owner group is available to support him, else you are taking a grave risk.
Deepak Anand

Clarification added 8 months ago:

The devices in Datacenter are like babies, you have to ensure to do a proper baby-sitting else you will end up hearing all the bawlings :-)

Cheers,
Deepak Anand

Sanjay Mohindroo, Technical Project Management professional

see all my answers

Hi
Administrators vs no. of servers depends on the stability of the organization and the stability of the OS being used. I have driven an initiative where I could drive the ratio from 150 per SA to 900 per SA. This involved a lot of standardization on the OS level as well as a focus effort to reduce known problems and working on eradication of the problems rather than break fix efforts. A lot of focus was also driven on the quality and training of the SA's.

Barry Lewis, CISSP,CISM, Owner, Cerberus ISC Inc., Information security specialist

see all my answers

Best Answers in: Information Security (2)... see more, Professional Networking (1), Enterprise Software (1), Computers and Software (1) see less

I think you might also find that part of the problem lies in what activities you perform as an server adminstrator and how well that is standardized and automated. For example, are you also performing security functions in an environment with a lot of turnover versus little turnover (or not doing security because others do that). How many applications and what sort of involvment might you have in changes etc.

One method might be to determine how many servers there are (so many companies have little idea it seems), then how many changes occur on a give time frame, how much of a backlog is there, and how many admins are there with a)lots of experience, and b)little experience. Obviously, experience will influence your count, perhaps significantly.

Recommended links



Etc

FAIR USE NOTICE This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available in our efforts to advance understanding of environmental, political, human rights, economic, democracy, scientific, and social justice issues, etc. We believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit exclusivly for research and educational purposes.   If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner. 

ABUSE: IPs or network segments from which we detect a stream of probes might be blocked for no less then 90 days. Multiple types of probes increase this period.  

Society

Groupthink : Two Party System as Polyarchy : Corruption of Regulators : Bureaucracies : Understanding Micromanagers and Control Freaks : Toxic Managers :   Harvard Mafia : Diplomatic Communication : Surviving a Bad Performance Review : Insufficient Retirement Funds as Immanent Problem of Neoliberal Regime : PseudoScience : Who Rules America : Neoliberalism  : The Iron Law of Oligarchy : Libertarian Philosophy

Quotes

War and Peace : Skeptical Finance : John Kenneth Galbraith :Talleyrand : Oscar Wilde : Otto Von Bismarck : Keynes : George Carlin : Skeptics : Propaganda  : SE quotes : Language Design and Programming Quotes : Random IT-related quotesSomerset Maugham : Marcus Aurelius : Kurt Vonnegut : Eric Hoffer : Winston Churchill : Napoleon Bonaparte : Ambrose BierceBernard Shaw : Mark Twain Quotes

Bulletin:

Vol 25, No.12 (December, 2013) Rational Fools vs. Efficient Crooks The efficient markets hypothesis : Political Skeptic Bulletin, 2013 : Unemployment Bulletin, 2010 :  Vol 23, No.10 (October, 2011) An observation about corporate security departments : Slightly Skeptical Euromaydan Chronicles, June 2014 : Greenspan legacy bulletin, 2008 : Vol 25, No.10 (October, 2013) Cryptolocker Trojan (Win32/Crilock.A) : Vol 25, No.08 (August, 2013) Cloud providers as intelligence collection hubs : Financial Humor Bulletin, 2010 : Inequality Bulletin, 2009 : Financial Humor Bulletin, 2008 : Copyleft Problems Bulletin, 2004 : Financial Humor Bulletin, 2011 : Energy Bulletin, 2010 : Malware Protection Bulletin, 2010 : Vol 26, No.1 (January, 2013) Object-Oriented Cult : Political Skeptic Bulletin, 2011 : Vol 23, No.11 (November, 2011) Softpanorama classification of sysadmin horror stories : Vol 25, No.05 (May, 2013) Corporate bullshit as a communication method  : Vol 25, No.06 (June, 2013) A Note on the Relationship of Brooks Law and Conway Law

History:

Fifty glorious years (1950-2000): the triumph of the US computer engineering : Donald Knuth : TAoCP and its Influence of Computer Science : Richard Stallman : Linus Torvalds  : Larry Wall  : John K. Ousterhout : CTSS : Multix OS Unix History : Unix shell history : VI editor : History of pipes concept : Solaris : MS DOSProgramming Languages History : PL/1 : Simula 67 : C : History of GCC developmentScripting Languages : Perl history   : OS History : Mail : DNS : SSH : CPU Instruction Sets : SPARC systems 1987-2006 : Norton Commander : Norton Utilities : Norton Ghost : Frontpage history : Malware Defense History : GNU Screen : OSS early history

Classic books:

The Peter Principle : Parkinson Law : 1984 : The Mythical Man-MonthHow to Solve It by George Polya : The Art of Computer Programming : The Elements of Programming Style : The Unix Haterís Handbook : The Jargon file : The True Believer : Programming Pearls : The Good Soldier Svejk : The Power Elite

Most popular humor pages:

Manifest of the Softpanorama IT Slacker Society : Ten Commandments of the IT Slackers Society : Computer Humor Collection : BSD Logo Story : The Cuckoo's Egg : IT Slang : C++ Humor : ARE YOU A BBS ADDICT? : The Perl Purity Test : Object oriented programmers of all nations : Financial Humor : Financial Humor Bulletin, 2008 : Financial Humor Bulletin, 2010 : The Most Comprehensive Collection of Editor-related Humor : Programming Language Humor : Goldman Sachs related humor : Greenspan humor : C Humor : Scripting Humor : Real Programmers Humor : Web Humor : GPL-related Humor : OFM Humor : Politically Incorrect Humor : IDS Humor : "Linux Sucks" Humor : Russian Musical Humor : Best Russian Programmer Humor : Microsoft plans to buy Catholic Church : Richard Stallman Related Humor : Admin Humor : Perl-related Humor : Linus Torvalds Related humor : PseudoScience Related Humor : Networking Humor : Shell Humor : Financial Humor Bulletin, 2011 : Financial Humor Bulletin, 2012 : Financial Humor Bulletin, 2013 : Java Humor : Software Engineering Humor : Sun Solaris Related Humor : Education Humor : IBM Humor : Assembler-related Humor : VIM Humor : Computer Viruses Humor : Bright tomorrow is rescheduled to a day after tomorrow : Classic Computer Humor

The Last but not Least


Copyright © 1996-2016 by Dr. Nikolai Bezroukov. www.softpanorama.org was created as a service to the UN Sustainable Development Networking Programme (SDNP) in the author free time. This document is an industrial compilation designed and created exclusively for educational use and is distributed under the Softpanorama Content License.

The site uses AdSense so you need to be aware of Google privacy policy. You you do not want to be tracked by Google please disable Javascript for this site. This site is perfectly usable without Javascript.

Original materials copyright belong to respective owners. Quotes are made for educational purposes only in compliance with the fair use doctrine.

FAIR USE NOTICE This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available to advance understanding of computer science, IT technology, economic, scientific, and social issues. We believe this constitutes a 'fair use' of any such copyrighted material as provided by section 107 of the US Copyright Law according to which such material can be distributed without profit exclusively for research and educational purposes.

This is a Spartan WHYFF (We Help You For Free) site written by people for whom English is not a native language. Grammar and spelling errors should be expected. The site contain some broken links as it develops like a living tree...

You can use PayPal to make a contribution, supporting development of this site and speed up access. In case softpanorama.org is down you can use the at softpanorama.info

Disclaimer:

The statements, views and opinions presented on this web page are those of the author (or referenced source) and are not endorsed by, nor do they necessarily reflect, the opinions of the author present and former employers, SDNP or any other organization the author may be associated with. We do not warrant the correctness of the information provided or its fitness for any purpose.

Last modified: September 12, 2017