Slashdot Mirror
How Hard is it to Manage Different Unices?
vrmlguy asks: "Where I work has several Unix-based servers, all running the same vendor's OS. We are getting ready to buy another big server, and management wants to get bids from other vendors. However, our staff is only familar with our current vendor's OS. Yes, I know that any two flavors of Unix are more alike than not, and yes, I know about the Rosetta Stone for Unix that makes it easy to transfer skills. I want to know about the down-side: What's the difference in the cost of operations between a mono-culture and a shop running two or more vendors' OSs?"
You have a team of mechanics, and for the last 20 years all they have serviced, as well as driven themselves are Ford automobiles. Now, your boss tells them to jump right in and service Chevrolet autos too. How easy will this change be? Depends on the mechanics and how they've been trained I suppose.
It has always been my opinion that if you have people who understand the concepts and underpinnings of how *nix systems work, the flavor of the OS doesn't matter. People who have a good understanding from an abstract point of view will easily pick the differences in syntax, location, etc.
"I'm just here to regulate funkiness."
I have been doing SysAdmin for a mere 3.5 years. At work we have a few that I need to deal with every so often: HP-UX, Solaris, AIX, SuSE, and Redhat. Seems like all the Techie SysAdmins (myself included) have taken to using the OS they know for a project and then expecting other folks to pick it up ("Oh, it is close enough..."). I wish our company has the insight to pick a standard - 1 or 2! I would LOVE for the versions to follow the same file placement conventions, command conventions, and system management tools. Maybe, someday....
"If you are on fire you can just stop, drop, and roll. If you fall into Lava you are just dead." - my 5yr old daughter
What's the difference in the cost of operations between a mono-culture and a shop running two or more vendors' OSs?
$32,593.12
Now can we stop with these stupid, inane questions? I would rather read Jon Katz than these awful Ask Slashdot questions of the past 3 months or so.
Your biggest expense is going to be training, but your company will probably choose to take it out of your clients' and employees' pockets by doing it "on-the-job". Next up would be licensing fees.
Unless Vendor B is offering a competitive upgrade from Vendor A's software, it would be much cheaper to negotiate an additional Server and Client license pack from your existing vendor than to enter into a new business relationship with some new vendor. Unless, of course, the new "vendor" is (sigh) Linux.
If you fall off a building, go real limp, because maybe you'll look like a dummy and people will be like hey, free dummy
For the users, all of the familiar commands shoud work fine. But maintaining the boxen will have a cost. For instance, I know how to create a disk partition under both Linux and AIX and can say that the process is totally different. Also, you'll have to keep two different platforms up to date with the latest patches. And don't forget your apps, which probably won't have binary compatibility. You'll have to make sure that all of the apps that you wish to run are ported to your new Unix flavor of choice.
"Avoid employing unlucky people - throw half of the pile of CVs in the bin without reading them." -- David Brent
Not a flippant post.
The quality of Unix sysadmins has declined so much over the past decade that what passes for a sysadmin right now is what I used to call "an operator".
We have 5 unix sysadmins (major transportation company). Not one of them could write a shell script if their life depended on it.
They insist on doing everything by hand and then complain there are no automated tools to them. Their definition of an automated tool really means "graphical front end to those grubby text commands".
They have no appreciation for the modularity of unix, and they look longingly at Windows servers.
Meanwhile, they're all getting paid twice what they're worth because apparently as dumb as the Unix sysadmins are, the NT ones are apparently on a different evolutionary scale where "rock" is considered the most intelligent life form.
So my point is that getting these sysadmins to switch won't happen. They'll piss, bitch and moan about the opportunity to learn something to enhance their skills, then complain the application is screwing up "their" servers.
If only ASPs would take off, my life would be much better, because sysadmin skills suck so bad, black holes pale in comparision to the event horizon of these so-called admins.
So, is it just me or does it bother you that the "Rosetta Stone" states "This custom drawing feature requires IE 5"?
ALL HAIL BRAK!!!
The same principle applies to natural and computer languages - the more you know, the better you understand the fundamentals.
:)
Sure, you might know how to do x,y,and z on your Solaris box, but once you understand how to do it also on RedHat and AIX, you'll understand much better how it works conceptually. Then when you get an HP box, it'll be pretty easy.
Of course, don't run killall on HP.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
It is fairly easy to transfer sills from one version of UNIX to another.
Plus, it is greate for the resume. When you get tired of this job, get fired, laid off, or transfered you will find it much easier to find another job.
Some of the differences between different versions of UNIX include:
BSD or AT&T based
Disk tools
Adminstrative interfaces and GUIs (SAM, SMIT, etc)
Startup / Shutdown scripts (rc.d vs init.d)
User management
Included tools ("top" is a big one)
Backup and recovery (hp includes fbackup / frestore)
X-Windows (CDE, VUE, etc.)
Some if the similarities include:
user land tools (ps, ls, find, etc)
Directory structures are slowing becoming the same
Generally, this is not difficult to do, as long as your admins understand the bases of UNIX. (Vendor-centric admins sometimes don't, as they get dependent on their vendor's tools.)
The problems can arise with:
1. vendor-centric admins who aren't willing to learn
2. different service contracts creating differing expectations of uptime between systems
3. added costs from maintaining multiple service contracts and training on multiple platforms
4. finger-pointing, if the systems interact
5. rewriting in-house tools which are needed on the new platform, but were not written generically before
6. 3rd party licensing costs may differ (if you are licensing the same product on both OSs)
7. dilution of expertise, since your admins will have to be more generalists (this is often overbalanced by the expansion of perspective in problem-solving that comes from broader experience)
Other than that, I can't think of anything off the top of my head which would make this hard. Generally, it is not a problem to do.
-- Two men say they're Jesus. One of them must be wrong. - Dire Straits
on how you are using each platform. The biggest problems I have seen deal with propietary features in the different Unices. For instance, I worked as a Solaris Admin using NIS+ and while it supported authenicating other Unices that could just use NIS, it don't work well. But that was a years ago.
.cshrc) that set your preferences on the different boxes using uname.
Things that help include creating branchs in your login scripts (.profile or
There is a good O'Rielly book called "Unix for Oracle DBAs" that is a really good cross Unix reference that you should consider picking up.
There are a lot of variables you haven't talked about, e.g.:
1) What sort of applications do you have running on these servers and how interoperable are they? Does it matter how interoperable they are?
2) And further, are those apps dependent on that vendor's Unix?
3) How much resistance is there from the staff to learn something new?
Assuming the above aren't a problem, then it shouldn't really matter. Go open-source and save a buck or two.
I watched C-beams glitter in the dark near the Tannhauser gate.
I'm going to try and not sound like a troll here. But this Ask Slashdot question seems complete rubbish.
Coukd the Slashdot folks be a little more discriminating in their choice of questions, please? The most entertaining/thought-provoking parts of this story, seem to be the idiot troll posts. This is hardly a thought-provoking or difficult question to answer/figure out with the most miniscule of job skill.
To answer this silly question:
The difference is: a lot, due to training/familiarization, support contracts, possible hardware differences, etc. DUH.
It works like this . . .
You learn one flavor of UNIX, get to know it inside & out. And because of the shop, the job market, whatever, you start working with another flavor. And it will look weird because it's different.
Sometimes the differences are due to developers' choices, sometimes they fix problems existent in the first flavor you've encountered, sometimes they cause problems you didn't have in the first flavor. And sometimes what's weird about this new flavor is because the guy who set the computer up botched things.
Also, the longer you know one flavor of UNIX, the more likely you are to call any new flavor you encounter ``braindead".
Except when it comes to SCO. Trust me on that one.
Geoff
I think I see a trend here. Maybe for them it really would be easier to muzzle the entire internet than to produce p
What's the difference in the cost of operations between a mono-culture and a shop running two or more vendors' OSs?
How much of a raise are you asking for?
-... ---
Sure, an environment with only one vendor's OS deployed is easier for the admins to handle. However, if a problem develops, that problem will affect EVERY SINGLE MACHINE you have. Don't lose sight of that in your zeal to minimize the admins' workloads.
Learn to spell: nickel, missile, lose, solely, amendment, speech, kernel, probably, ridiculous, deity, hierarchy, versus
Also, running your software on these different flavors of UNIX will shake out some bugs that you might not find on just one or two platforms.
One major consideration will be the service contract with the vendors. With 2 vendors, you'll buy 2 service contracts. You can have all the best sys admins you want, but I'm sure you'll need at least a minimal service contract for a commercial Unix. Can adding a second contract add a lot to overall cost? Hard to say without any details of the company, but I'd guess adding a basic service contract from the new vendor will significantly add to TCO.
Developers: We can use your help.
Are you daft?
Open Source is a philosophy of software distribution, not a standard for setup and maintenance.
Try switching from Mandrake to SuSE without pulling out a few hairs relearning where all the init scripts are kept and how the system is configured and maintained. Then jump to a BSD for shits and giggles.
Are what's going to kill you. Having to support software and software interoperability between different platforms can be a serious pain. A mono-culture is easier when dealing with software. However if you are presented with a significant enough savings from another platform, consider it.
Your admins, if they're any good, should be able to adapt to a different UNIX easily. Yes there are differences, but not ones that should trouble an experienced admin any longer then it takes him to read a couple man pages.
If you are happy with the vendor you are with, and everyone likes working on their product, it makes NO sense to switch vendors to save a couple bucks. The technicians will spend (as read by management: waste) their time learning the ins and outs, do's and dont's of the new OS. Not to mention possible incompatibilities, and more wasted time futzing with network integration, plus warranty and support calls (Sun: Its your IBM box at fault! IBM: its your Sun box at fault!) If you are happy with the platform you are on, stay with it!
Explain that to management, and I'd be very suprised if they didnt continue with the origional vendor.
Statement of Bias: I "administer" several UNIX OS versions (Solaris, IRIX, Linux, occasional HP-UX), but in an isolated network with no outside connections (so very little emphasis on security).
Two factors come to mind:
No matter how close the systems are, you will still "loose" time to training (either formal or OJT) requirements for the new system. This may actually be a benefit for your staff (wider perspective, more to put on Resume).
Depending on how much focus is placed on security, you may end up doubling the time required to track vulnerabilities and install patches. Again, this may be an advantage as well since a single-os shop tends to have equal vulnerabilities on all systems. In a multi-os shop an attacker will have to work harder to get control of everything.
It is easist to manage servers from only one vender. Unix makes ti easy to transfer skills, but here is the contradiction: It is easier to manage servers from many different venders and versions, than to manage just one server that is different from the rest.
When you have all OSes the same it is easy because everything automaticly transfers. When all are different it is harder because you always have to remember the correct incarnation of each procedure, but because they are all different you get in the habbit of looking it up each time. When all are the same except for one machine you forget on that one machine that everything is different and you aply the wrong incarnation (ofte with disasterious results, see discussions of killall linux vs hpux on comp.risks) Because of this, the one different machine will get [invalid] complaints often due to these differences.
If you can't stick with one vender, then you should go with many so you are in the habbit of checking the differences. At the very least get some linux (debian, redhat, suse), and bsd (free, open, net) machines in house now, and use them for production. You need to make sure that your admins are used to subtile differences. The other alternative is to just stick with one vender, but not only do you pay more, but your admins become lower quality as they learn only one system. (think of it is a resume builder, you want different systems on your resume!)
It seems to me that the biggest cost is in sysadmin time. I figure it this way, at work I use a few UNIX systems. We have one machine running IRIX, a couple running BSD and one running Linux. Now, when I write a script one one of the BSD machines, it works on all of them, but it may or may not work on the Linux machine, and certainly won't work on the System V-esque IRIX machine.
Now, if your sysadmins employ a lot of scripts, figure you'll have to spend twice the time maintaining them if you have two different platforms that are not fully compatible. You can minimize this if you stick to POSIXly correct scripting, but you'll never completely eliminate it.
The same goes for custom programming. For instance, if you're running everything on BSD, and you want to take on a Sun machine running solaris, there may be some issues with the occasional socket call that Sun implments differently from the rest of the world.
So, the more custom scripting/custom apps you have, the more time your sysadmins will have to spend maintaining/porting/testing the stuff.
---
Play Six Pack Man. I
From personal experience, I found that backup and recovery can be quite different depending on the flavours involved. For example, I back up my AIX systems with /usr/bin/mksysb, ftp the file to a system that is connected to a tape library and copy the image to a 4mm tape. I can do a bare-metal recovery from that tape to any equivalent or better RS/6000 in about an hour or so and have an exact clone of the original system as of the date the backup was taken. In this regard, AIX rocks.
/usr/sbin/ufsdump, but restoring a system from tape is more involved, and I cannot restore that tape on a different class of Sun hardware.
My Solaris backup and recovery strategy is not as elegant in that I make backup tapes via
I do not expect both to work identically, but there are some significant differences between the two.
*** Where are we going? And what's with this handbasket?
The short term costs to retrain staff for the new system will be higher but the long term benefits will definitely outweigh them. Once you build a multi-OS capable IT department the cost to add new hardware later on becomes significantly less. By not being locked into one OS (and one vendor) you have the freedom and flexibility to choose the best solutions for future problems (as well as hunt for the lowest cost). The smart thing to do is diversify your IT shop as much as possible so that you can insure you always have the right tool for the right job. No single vendor or OS can provide all the answers, regardless of what IBM/Sun/Microsoft may try to tell you.
A more serious answer.
You have a skewed view of the UNIX admin world. In my workplace (small investment firm), we have 6 UNIX admins who jump at the chance to learn something new. They'll dive into Linux, then realize OpenBSD is better for their task because of its security and inhale the documentation, all while keeping a fleet of Solaris servers running for production work.
If one of them does not know how to program, he picks up a book and starts writing python in a couple of days.
It sounds like you and I are at the extreme ends of the UNIX admin experience, because my situation sounds so opposite to yours.
To the original poster: what kind of workplace is yours? If your UNIX people jump at new stuff, they'll soon figure out if the new system can be successfully integrated and how long it will take.
Just keep text-file logbooks as you learn new things about the different UNIX implementation. Keep them in a hierarchical database on an NFS file system or web server somewhere, name the directories and files consistently by OS and topic (topics such as DNS, network booting, firmware, SCSI naming conventions, package management, etc.).
I do this at home to juggle Solaris, OpenBSD, and Linux and it works well. If I forget how to setup DNS under Solaris, I just go to <base_dir>/Solaris/8/DNS_Setup.txt, for example.
Also, install all of the on-line documentation you can and have it network-accessible. For example, when the man pages aren't detailed enough, on-line Solaris Answerbooks can save the day.
Also, keep well-organized bookmark lists for useful websites, such as http://docs.sun.com or http://sunsolve.sun.com, that cover your particular UNIX.
Having any number of UNIX implementations really isn't unmanagable (unless they have broken network protocol implementations). The key to success is documentation and more documentation (and unambiguously sharp sysadmins). On that note, if you don't have faith in your system and network administrators, you should just give up and stick with one OS, since no amount of documentation helps a truly stupid human.
Healthcare article at Kuro5hin
Money.
You obviously already know that managing support contracts from multiple vendors is going to suck. I would also recommend taking a long hard look at ongoing support charges.
For example, we have both HP/UX and Sun platforms where I work. We have both servers and workstations. For the workstation support contracts on similarly sized machines, there was a world of difference in cost.
The annual fee for an HP C240 workstation was somewhere between $2500 and $3000. The same annual charge for a Sun Ultra of equal speed, was between $1000 and $1500. Multiply that by the number of workstaions you have to maintain, and it can add up very quickly.
The up front cost typically isn't where they get you. It's on the back end. I would research the back end on all of the platforms you are considering very carefully before making any final decisions.
Hope that helps a little.
But today, if you give good training to your staff and give them good management tool, like the Tivoli suite from IBM
I have a medium-sized Tivoli installation, and it has not really reduced the number of SysAdmins we have. Tivoli still has some bugs in it, and some of the modules (i.e. software distribution) do not work consistently. Furthermore, Tivoli has problem running large scripts on remote servers.
In fairness, it does a very good job of monitoring systems, and it takes care of the more mind-numbingly dull and repetitive tasks for us, but we still need to have skilled UNIX admins around.
*** Where are we going? And what's with this handbasket?
For example, on Solaris (without Veritas Volume Manager), you have to "carve out" your disk filesystem by filesystem, and work with devices in /dev/dsk/cAtBdCsD format. On AIX, the concept is totally different with Logical Volume Manager, wherein filesystems can be created on the fly. But HP-UX uses both in an odd fashion, forsaking slices and using a "castrated" form of LVM. This is just one example, as you will find other things in HP-UX such as the useradd command being identical to Linux and Solaris, and the SAM tool being very close to AIX's SMIT utility.
In the end, as you will find, there is no uber-Unix that will carry over to all of the other flavors. IMHO, HP-UX is as close as you will get. But, my personal preference of all Unices is AIX due to its ease of use (an IBM tool easy to use? I know it sounds like an oxymoron) and robust capabilities, combined with Linux integration in the most recent versions. Flame as you will, I'm interested in hearing anybody else's insight.
--Chag
Apparently the Rosetta Stone can survive 4,000 years of Mother Nature's worst, but cripples in minutes under the power of the Slashdot effect.
Most companies I have worked at or know people at go with a third party backup solution such as the ones from Tivoli or Veritas.
Makes your backup/recovery fairly consistent across different products, plus everything can then be managed from a central console.
I've found that operationally, my ability to move between Tru64, Soalris, AIX, HP-UX is relatively seamless. My biggest hurdles have come when doing hardware troubleshooting, upgrades and maintenance. Each vendor has their own unique approach to device names, hardware settings and architectures, which I've found to be the most difficult to master when moving between unixes.
Guess folks at SGI never heard of the "path of least astonishment", such as printing a usage message if there's no arguments. Then again you could argue that it's not very astonishing if typing 'killall' really does kill *all*.
Three dits, four dits, two dits, dah!
Radio, radio, rah rah rah!
"What kind of unixes do you run?"
"Oh, we have both kinds. RedHat and Debian.
The reason homogenous environments are easier to manage than heterogenous environments is due to complexity.
Simply put, if every server and workstation is identical, interoperability is not an issue, and the work associated with tracking, testing, and applying changes to that one, homogenous OS image is minimal.
The moment you branch out into different configuations of the same OS version, different OS versions, or different OS platforms, you've increased the complexity of the system, and thus increased your workload. Suddenly, interoperability is a factor in every decision, and issues with multiple versions and/or vendors must be tracked.
I've been meaning to write a short paper on this for some time, and attempt to relate it to Christopher Langton's Lamba parameter for the measurement of complexity (in the 3rd Annual Proceedings of the Artificial Life Conference). I've studied the identification of single points of failure for some time, as well as the question of "how many sysadmins do I need?". Both answers are directly related to the complexity of the system being managed (here, defining "system" as the collection of applications, OSes, hardware, and networks that comprise the scope of a sysadmin's responsibility). There are indeed identifiable factors that define the heterogeneity of an environment, and the ways in which these dimensions impact such things as the number of SA's required to manage them can be defined.
.@.
Here is a Google cache of the link he mentioned
But not as hard as you'd imagine. I use AIX, Solaris, and HP-UX on a weekly basis, as well as MacOS, MacOSX, and w2k (workstation and server).
/opt/apps/local
The biggest problem with a mixed environment is keeping it up to date with patches etc. Keeping track of that stuff is a complete PITA; I can't imagine how much more difficult it is in Linux, where the patches aren't on the vendor site (are they?).
Besides that, the big thing that you'll need to do is make sure everything is sort of in the same directory structure. For apps that you install, put them in the -exact- same directory. For example, all my unix boxes have the same layout:
/opt/apps
/opt/servers
/opt/data
/opt/src
/ usr/local ->
That way, it doesn't matter as much which box I'm on, and I don't have to remember exceptional cases. It also makes maintenance easier, because all the exciting stuff (non base operating system) is in a known structure. That means you can write scripts, etc to monitor everything and you don't have to change them on a per-host basis. It also means you can just copy the config.status from box to box (or directory to directory) and build without reconfiguring everything.
'Luck!
http://www-ccar.colorado.edu/~jasp2/Graph.html http://cam.radioactivecat.com/unix-rosetta.pdf Not as graphically friendly as the orginal... But, Still gets the point across...
how about 2400 machines run by 2 operators, 2 helpdesk-types, and 3 admins.
Using what you ask? Korn Shell code.
this includes automated, unattended reinstalls, backups, printer selection and setup, software installs, X configs. Everything, and we were bored.
So Tivoli, or Unicenter? yeah, it's doable with them, but it's also doable with shell scripts. And a hell of a lot cheaper.
"We are not tolerant people. We prefer drastically effective solutions"
1) You can install the same shell on just about all UNIX's. Most people where I am prefer tcsh as it has some nice features.
2) You can standardize on scripts, either use csh (blah) or sh. We prefer sh as it is found on just about EVERY unix (Sun, HP, AIX, BSD's, Linux).
3) Avoid vender extensions to the basic shell. HP has done some aweful things there in its bourne shell and they are not compatible with Sun and in some cases Linux either. I.E. Always use `cat foo` and not $(cat foo) in sh scripts. There are other things like that.
There are problems in supporting more than one UNIX, but there are also workarounds if you do it right.
Only 'flamers' flame!
NT isn't quite that easy. Curveballs are common and if your NT admins can't adapt, things are going to be somewhat painful. In my experience, the NT admins that you would want to keep around are also the sorts that would also be able to jump into Unix admin with few problems.
Having a mixed shop might actually allow you to quickly determine who the wannabes are versus your real talent.
Allowing the wannabes to plod along happily day to day may cost you less in the short run. However, it will cost you more as soon as you occur problems of any significant complexity.
A Pirate and a Puritan look the same on a balance sheet.
I don't know why, but when I read your post, I immediately thought of this thing.
Although it looks like a complete joke, there is a lot of truth in there.
echo '[q]sa[ln0=aln80~Psnlbx]16isb572CCB9AE9DB03273snlbxq' |dc
Hello,
I've been sysadmin at various points for
a small cluster which has had up to 6
different UNIXes:
Digital UNIX
HP-UX
Linux
SunOS 4.1.x
Ultrix
Irix
Now, I was able to manage each of these pretty
OK, Unixes *are* alike. However, getting patches
and whatnot differs over each arch.
So crudely, I would say:
SysadminWork =
A * number of UNIXES
+ sum_i(Bi * number of machines_i)
where A is a very big constant,
i is the index of each UNIX,
Bi is a small constant, the marginal extra
effort to maintain one more machine of type i
What I mean is this:
for each UNIX, you have to do a fairly large
amount of research + effort to learn/aquire
materials and knowledge for things like upgrades.
Having done that, it's easy for you to maintain
another UNIX box of that type: the cost of each
extra machine is low, and you can do things
efficiently via scripts.
So the least wasteful way to use your sysadmin
is to have one arch/OS.
In my case, my life became progressively easier
as I got rid of UNIXes and concentrated on running
Linux only.
I mean, waht do you do? Do you just serve up content on software someone else writes? (Http, or SQL database?)
Or do you write your own real-time communication software? Writing device drivers across platforms can be sticky (if you are writing a high level device driver, utlizing CDLI or DLPI) to down right icky (you go down to the metal).
For us, switching platforms has a higher cost than the money spent on the boxen.
In the future, I would want to not be isolated from my friends in the Space Station.
> You'll call SCO 'braindead' no matter what.
Exactly my point.
What else can you say about a UNIX flavor developed by Microsoft? It takes all of the user unfriendliness of UNIX & combines it with the bad programming habits of MS. And SCO (before they were bought out by Caldera) failed horribly at maintaining the resulting code.
At least Caldera did the sensible thing: let SCO die, & offered all of the customers still using it a way to migrate to a better OS.
Geoff
I think I see a trend here. Maybe for them it really would be easier to muzzle the entire internet than to produce p
3) Avoid vender extensions to the basic shell. HP has done some aweful things there in its bourne shell and they are not compatible with Sun and in some cases Linux either. I.E. Always use `cat foo` and not $(cat foo) in sh scripts. There are other things like that.
...
...
... file
/.
...
... /etc/passwd, normal find /tmp -print would output
go here...
And of course, if you've been following along for a week or two, you know that this (BING!) is a Useless Use of Cat!
Rememeber, nearly all cases where you have:
cat file | some_command and its args
you can rewrite it as:
<file some_command and its args
and in some cases, such as this one, you can move the filename to the arglist as in:
some_command and its args
Just another Useless Use of
Dangerous Backticks
A special idiom to pay attention to, because it's basically always wrong, is this:
for f in `cat file`; do
done
Apart from the classical Useless Use of Cat here, the backticks are outright dangerous, unless you know the result of the backticks is going to be less than or equal to how long a command line your shell can accept. (Actually, this is a kernel limitation. The constant ARG_MAX in your limits.h should tell you how much your own system can take. POSIX requires ARG_MAX to be at least 4,096 bytes.)
Incidentally, this is also one of the Very Ancient Recurring Threads in comp.unix.shell so don't make the mistake of posting anything that resembles this.
The right way to do this is
while read f; do
done
/tmp/moo
/etc/passwd
and xargs would see two file names here. Changing the record separator to ASCII 0 means it's now valid for a file name to span multiple lines, so this becomes a non-issue.
As far as problems go, the only real problem you have is getting used to the new environment. My company is running HP-UX 11, AIX and RedHat Linux. HP-UX is a dream to configure and when I have to work on AIX, I have to most of the time, take the back roads through the console. It can be a pain but it's just like any new system. You just have to learn it. Oh, and sometimes root on HP-UX != root on AIX... but we're working on it.
I have used 5 different UNICES for over 12 years, and a majority of them at the same time. Solaris and AIX were the 23 biggies. If you "think" in UNIX, you'l realize that they are all the same. Although AIX is the most non-UNIX of the UNICES (they chnage the name of commands just to be different), but as long as you have a UNIX frame of mind, there is no learning curve. I prefer different UNIXES. I never liked being too dependant on one vendor. Man, as long as you have your scripts in order (you DO, don't you), and made them with a UNIX mentality, you'll have few problems. I personally don't know any UNIX guy who only knows ONE UNIX. That's odd. My last job, we had RS/6000's for almost everything, but Sun for Oracle (since it ran better on Sun). How did I know that? We got an HP box, a Sun Box, a Compaq Alpha running Tru64, and an RS/6000 and performed tests on each of them. Everr hear of IQ and OQ? Integration Quality and Operation Quality. What BEST supports the needs of the client? A real UNIX admin learns to think outside the box. Hiring someone who is a "one platform man" is detrimental. Or someone who refuses to learn another platform. That's the mentality of an NT admin, not a UNIX admin. I've always taken a consultants viewpoint. Think of ROI. Think of quality. Think of uptime. It all comes down to money. If a 100 person company has 14 RS/6000 M80's, SOMEBODY is getting fucked over! Sure, one vendor makes life easier for the accounting dept. But YOU have a job to do. And your job is to ultimately keep your company running. As UNIX admins, we are a very precious commodity that the dot bomb crash has made every body forget. We are supposed to be the example of computer usage. Is it COOL to have 30 servers, or do we realy NEED 30 servers? SInce IT is usually the dept w/ the highest costs, we save face when we choose the BEST tool for the job. If you have more than 2 FUL-TIME admins in a 100 person company, something is WRONG. REALLY wrong. (Not counting integration and building) I've been at companies where I was the only admin for a 200 server farm. Call the consultant when things got hairy. But all in all, UNIX is supposed to WORK. If it's not, there is a problem
If you're not a Liberal in your 20's, then you have no heart.If you're still a Liberal in your 30's you have no brain.
Note that training cost should be added to the overall cost of the equipment purchase. Tell your managers to send the staff to a training course offered by the vendor if they do buy this new box with a different OS.
Amen to this statement! I've discovered some pretty obscure bugs when switching between OSF 4, OSF 5.1 and Linux. For each change of OS new "undocumented features" were discovered.
Windows XP is SOOOO not unix that it's not even funny. I wish I had what you were smoking because it must be pretty good.
According to this, "The POSIX subsystem included with Windows NT and Windows 2000 is not included with Windows XP Professional." It is a separate ($$$) product called Windows Services for UNIX. Don't know if this applies to the "server" flavor of XP though. MS can be so inconsistant for a monopoly... I like how they call it "windows services for unix" when it's really "UNIX compatability for Windows." Ahh, MS marketing...
Besides the "optional" posix API layer (and optional generic utilities,) there is NOTHING unix like in Windows.
we had multiple OSes to support.
We didn't have the luxury of 2, no . . . we had somewhere between 10 and 20 different versions of operating systems, that is if you include different revisions, etc.
we had everything from SCO to Solaris, to NT 4 to 2k, it was nasty.
The company had bought out a bunch of little ISPs and just threw all their boxes in the racks and made us try and get em all on the network.
Many of these were bought out ISPs and the admins were fired, so of course half of em had no passwords, and a bunch had all kinds of nasty little quirks.
I would say stick with no more than 2 versions at a time, maybe 3.
Different distros have their strong points and weak points, so balance it that way. There is not much of a learning curve unless you have like Solaris and Redhat and BSD in the same building.
Then you start forgetting which system commands work where because you log into em so frequently to do different things.
Its really not an issue of learning curve, its more an issue of annoyance.
the best recommendation I have is to make scripts to do the simplest tasks, that made things so much easier for us in our situation.
Where I started, we had five main types of unix running (SunOS, HP-UX, OSF/1, AIX, and Linux) and multiple revs of all of them. It was trying at times, sure, but if the admin team is able and willing to think about things and isn't afraid to make mistakes, they will get the job done.
Easilly the number one cost is admin time to learn all the tricks necessary to get the job done.
you should read everything on the internet as if it had "but I'm probably talking out of my ass" appended to it.
Lots of people have mentioned the various sources of additional costs that can come from having a multi-vendor system environment.
There are potential savings as well:
1. Better rounded employees will be better able to assess the benifits subtle differences in technologies for different applications.
2. Ability to attract good people with established skills on one platform who want the chance to transfer those skills to another platform.
3. *ability* to scare vendors when necessary into giving you a better deal because you already have in house expertise on competing systems. This can be very valuable when negotiating upgrades, new systems and renewed maintainance contracts. Just be careful when and how you wield it. If you are too heavy handed, they may just decide that they are going to loose you and try to milk all they can from you during the transition. It is probably best as a subtle threat wielded when trying to do a deal with them at the end of a tight quarter when their sales team is driven more by tactics to maximize short term revenue at the possible expense of strategic influences on pricing.
Going from SuSE to LFS wasn't as bad as you might think. The main difference that I can recall is that the scripts that control various services live in /sbin/init.d on a SuSE box, but /etc/init.d on an LFS box.
The biggest difficulty is dealing with the automated config software that most distros use. I can usually set up most things on a SuSE box through YaST, but I haven't figured out whatever config utilities are used by the one Redh*t box at work that I haven't nuked yet. (Then again, I ran SuSE at home for a couple of years. I ran Slackware before that, and SLS before that. I've never installed Redh*t or had to deal with it prior to my current job.) I'd still rather tweak the different config files manually for the few apps that need adjustment, though; it's usually easier to dial in the exact setup you want that way. That's why most of the Linux machines I control run LFS now (the only exceptions are the aforementioned Redh*t box and an ancient 486 print server that was set up with Slackware because I didn't want to wait for that slug to build LFS).
20 January 2017: the End of an Error.
I am in the process of rebuilding many of our "legacy" SGI, Apple, and even older Sun systems with Debian. Fewer security holes, homogenous and simple to manage (especially with apt), fast and lightweight - and runs on practically everything we have.
Basically, just choose what you feel is the best server offering (because of price, construction quality, hardware support, and track record), and once the initial install is done, no one knows the difference. :)
A government is a body of people notably ungoverned - AC
It's NEVER cheaper to be multi-platform. More flexible, yes. Right platform for the app, yes again. Cheaper, no. The overhead of managing multiple flavors is large. It matters not who you have for sysadmins, or how capable they are. Patch management, change control, binary compatability, backups, security management, OS upgrades, service contracts, hardware compatability, etc. are all issues that cost you more. You end up having to do the same work over and over for each flavor. Been there, done that.
A TRIVIAL example would be changing your IP address space. Each flavor maintains it config in a different way. It doesn't matter that you know how each one differs. You won't be able to write a simple script that just makes the changes (it would be a complex script if you even chose to do it via script. Or you would write a script for each platform. You would probably end up doing it by hand.)
Another trivial example would be initial system load. With solaris for example, you setup (and maintain) a jumpstart server. When you get a new machine, an hour later you have a complete environment setup with all your customizations, up to date patches, etc. without hardly lifting a finger. Now add IRIX, Redhat, debian, freebsd, aix, HP, OSX into the mix. See the problem?
The list of examples of all the additional costs associated with maintaining multiple flavors is virtually endless.
You basically have it backwards. It's cheaper in the SHORT run. You can shop based on price. Initial setup isn't that bad. It's the LONG term maintenance costs that get you. It's ALWAYS easier / faster / cheaper to only have one platform to maintain.
Knowing that (as an example) AIX has a pretty self tuning kernel, that Solaris has a modular kernel, and that UnixWare needs a recompile (relink) for any minor changes forces the admin to think about the operating system instead of just drooling on the keyboard.
The biggest differences are still SysV vs BSD. Understanding those is vital in a mixed OS environment. Beyond that, there are usually differences in disk layout (and filesystems), but they just add to the rich diversity that is my favourite OS.
At my work, we are big users of Solaris, but because we develop software for multiple platforms, I've also had exposure to AIX, UnixWare, Sequent Dynix/PTS, HP-UX and DRS/NX. These days we've dropped Dynix/PTS (EOL anyway I think), HP-UX (too expensive for our customers), and DRS/NX (dead?) but we're looking to port to OpenUnix 8 and Red Hat Linux, so things are still pretty mixed. I just think it's a shame that I don't get to work with HP-UX and that Unixware is dying (yes - I like it!).
We also port to NT/2000, which is good to compare - it's a nightmare to work with when used to UNIX.
You don't give much background information. If you are talking about a shopt hat has one or two mid-level solaris boxes and is looking at a new server for like, some application you run.. sure, look elsewhere.
IF you are talking bout BIG servers, and some heavy investment into infrastructure..
Let's say the only change you should make is to another vendor for the future.
It's not about whether not you can administer one box or the other... it's about simplicity. Don't make things more complicated than they have to be.
We have Linux and HPUX (used to also have Irix and Solaris). No real problem for the most part.
Usually what I find myself thinking is "why can't I do this/that on HPUX". Until we went to HPUX 11, the only thing HPUX used PAM for was CDE - now the whole system uses it. They've added something called cifsclient, which does the same as smbmount. You can even install ipfilter's into the OS now.
So using the actual OS isn't that bad - a few differences in the way some things are done here and there. Hardware can be night and day though.
I'm saying it's more expensive in the short run because I'm assuming they take the time to set up the scripting and automanagement at the beginning. Once they set it up adding boxes from the same vender is simple. But yes, I will agree (as you pointed out) for some problems having multiple OSes can be a pain. I would submit though that being locked into one vender can be just as costly and painful. Look at the whole Microsoft debacle when they decided to change their licensing and strong armed billions of dollars out of Microsoft only shops.
First thing I do when I start administering Unices is to install GNU utilities. At least this will lessen your headache.
Its simple. If you standardize across a platform, any programming or maintenance work becomes just that much easier. Need to write a new fubar script? Well, would you rather write it for 1 platform or 10? Would you rather debug 1 platform or 10? etc... etc... etc...
/. questions?
Can't we add an age/experience limit to the ask
Anonymous posts are filtered.
I tailored my post to the smaller company as I assumed this was. Cost metrics change when you are fortune 1000.
In the huge corp environment, you are dealing with huge quantities of boxes and people, and the cost differential between vendors starts to make a difference. In this case, you generally buy the best box (os) for the application.
I'd also like to ammend my response somewhat. While I still feel that it's more cost effective to stick with one vendor, there are exceptions. Pick an open-source flavor as your general purpose unix box, such as Debian linux or FreeBSD, etc., but only one flavor. They make great print / file servers, name servers, mail relays, intranet servers, proxies, administrative boxes, whatever. Pick a good intel box, and stick with it (compaq's are not too bad. DL360's are great little boxes, and hardware raid mirroring is handy.) You want to be able to move parts around.
I REALLY like to have fairly single purpose boxes. I lock them way down, and only install the absolute minimum needed for the particular function (which is why I like debian over redhat in a server role.)
Basically, bang for the buck you don't want to spend "Sun" money on a name server / mail relay. That's kinda silly. Yeah, sun has some new low-end boxes, but the value just isn't there.
My personal preference has been Sun for big iron / complex commercial apps and debian/intel for my everday little stuff.
The big point is that you don't want to end up down the road supporting 7 flavors of Unix. That really sucks. It's bad enough supporting multiple versions of the OS from the same vendor (which you frequently need to do because of app issues.)
Identical O/S and a super powerful oo scripting language everywhere for stiching the apps together.
Only half in jest, but it's a viable solution none the less.
it's not the tools: they're all GNU, and as you know, GNU's not Unix.
is it just something that implements read/write/fork/exec - a kernel? well, cygwin implements those.
& maybe a different phone number for ordering parts.
A mechanic that can't cope with working on different vehicles wouldn't even pass their apprenticeship.
I wouldn't even find a real Ferrari challenge over a Holden of Ford Falcon of the same period. Mind you it would take a lot longer, complexion just adds more time.
Really doing a clutch on a old HQ Holden with a 3 on the tree manual isn't much different than doing one on a slotted 5 speed Hewland transaxle of a Ferrari of the same period, say a Dino, except that there's a lot more bolts & screws to undo & redo to get to the clutch.
Hmmmm. It's a little problematic that you only have one flavor of Unix now.
My experience is that managing different flavors of Unix is a lot like porting to different flavors of Unix.
That is, learning to go from the first to the second platform is often the most difficult step. Then going from the second to the third is slightly easier, etc.
Level of difficulty = 1 - exp(-n)
Thus, for your situtation, the difficulty of migration is going to be directly related to how similar the 2nd platform is to the first.
In the old days it was easy to distinguish the BSD flavors from the SysV flavors as a significant learning hump that had to be overcome.
"Provided by the management for your protection."
For drivers you can split the driver in two peices. The part of the driver that interacts with the kernel / OS. This should have about 5 functions: READ, WRITE, STATUS, OPEN, CLOSE. Then you can write the second part of the driver to call these generic functions putting the absolute minimum in the functions, but also putting the parts that are OS specific there as well.
Only 'flamers' flame!
As a senior systems engineer from a similar organization (Carrier1 (FALCO!)) I can say there were no issue running a multi unix environment, and I've never had any issue with it at any of my previous companies (nor have any of the engineers I've worked with).
:). I had Mac OS X, GNU/HURD, Debian and Solaris all on my desk at one point.
:)
At Carrier1 had FreeBSD, Red Hat & Debian Linux, Solaris 9 & 9, HP-UX, even GNU/Hurd and Mac OS X (well, on *my* system
The only problem I've ever had is the fairly trivial (?!) one of getting the command flags right - stuff like the 'ps','route','ipchains, 'ipfw' and 'ifconfig' commands syntax being different, the different flags for package management tools, that sort of thing.
I quickly came to realise that it's not possible to remember all the flags for all programs and remember the best way to do something on a particular system if you are busy all the time, things just seem to seep out. This happens if you are spending lots of time programming or in meetings or working on large projects - in which case you might not touch one type of system for months (until there is a problem with it), at which point you find your self quickly reading man pages and referring to Google a lot. All you need to do is remeber what's improrant, especially things you'll need for troubleshooting, and not worry about the rest - it's enough to know about tool's like Solaris 'ndd' and Linux's 'mknod' and what they do, if you need to remeber exactly how to use them in a given instance you can refer to man pages, O'Reilly Books or Google (which I often find the fastest).
Staying current, reading Freshmeat everyday, installing and configuring new Unixes and new & un-familer packages regularly, being on mailing lists and reading Slashdot are good ways to stay up to date - the more you know the less likely you are to run into something completely unexpected. If your resourceful (which you should be as a Systems Engineer) the only real problems arise went you don't even know where to start, everything else is a piece of cake.
Basically, if you really know unix (and are not just a Red Hat Linux or Solaris flunky who has convinced themselves they are Gurus while they still run Windows 2000 day to day) then you won't have any problems.
Oh, and making lame excuses like 'well I need Windows for work stuff' and 'they won't let me run Unix on my desktop' DO NOT wash - they are just that - excuses for lameness.
I have been for job interviews and been introduced to guys who called themselves (literally!) 'Unix Gods', yet they had only ever used Solaris - if you have any of those you are in deep shit right now. [ Needless to say I ran a mile! ]
Most people fall somewhere in the middle of those two, you'll probably only have one or two decent guys, if your lucky, though if you need to ask you are very possibly in trouble already!
YMMV.
At our site we have HP-UX, OpenVMS, AIX, Solaris, and probably one or two I don't know about that our Midrange Systems people manage. There are three of them to service 15 or so major 24x7 medical-related systems. They can do it because they are darn good, we tend to dump the apps support on the apps vendor by contract, and the management doesn't interfere or hobble them with various insanities. Of the three the good admin one is most important, although having good management probably helps retain the good techs.
They say they like AIX best due to the ease of admin.
________________________________________ History Must Not Fall Into The Wrong Hands ___________________________________
IMHO, the single most significant difference between different vendors' offerings is the difference in what happens before you get to the OS. If you have a dead machine and want to run firmware diagnostics, or boot a different OS image or whatever, then the different platforms are almost completely proprietary.
There's the Intel stack - with the MBR outside the disc partitions, and the Bios limitations with 1024 cylinders, and the need for a tiny LILO or GRUB to get stuff from the filesystem.
There's the Sparc stack - with the boot blocks inside the disc partitions and Firmware capable of booting from the network.
I had brief encounters with the HP K-series stack once, and had to bluff my way through somehow.
While the differences in installers, and patchers, and package management are all significant, the ability to use the firmware layer effectively when the machine won't boot normally and the PHB is looking over your shoulder is the main difference between platforms.
Dunstan
The last scintilla of doubt just rode out of town
Also, the longer you know one flavor of UNIX, the more likely you are to call any new flavor you encounter ``braindead".
Except when it comes to SCO. Trust me on that one.
The longer you know one flavor of UNIX, specifically SCO,
then you are NOT likely to call ANY new flavor "braindead".
Is SCO the UNIX that Microsoft/UNISYS has the way out of?
Is NT (or whatever the current version is) "braindead" compared to SCO?