It's probably another very big downfall. A server hosting AD or group policy isn't touched by end users therefore you'd think it would be a much bigger possibility to run and open source OS on it if the features were available. And since it tends to be more secure may even be an attractive option.
It's probably another very big downfall. A server hosting AD or group policy isn't touched by end users therefore you'd think it would be a much bigger possibility to run and open source OS on it if the features were available. And since it tends to be more secure may even be an attractive option.
sourceninja
Diamond Member
- Mar 8, 2005
- 8,805
- 65
- 91
This is basically what I have done at my company. When I got here it was 100% microsoft on the desktop, 100% novell for the desktop network, 100% Solaris for the backend servers. As of the end of this year we will have moved to a linux server backend, positioned ourselves to replace novell (still not sure how we want to go on this), and have migrated to linux on some desktops (kiosks, etc) and open office for a majority of the staff.
I've also used virtualization to save us large amounts of money and at the same time giving us the flexibility to change rapidly instead of our old model of no big changes except for once a year. The biggest cost cut was solaris to linux. Instead of running solaris on sparc hardware and paying 10,000.00 a server. We are running linux on intel hardware and paying 3000.00 a server. With our new virtualization project we are going to reduce that cost even more. I'm installing a blade system that will handle our entire physical server environment + some for the same cost it would take to replace our aging 5 year old servers. Obviously we will still have physical servers, but ESX + linux has reduce the overhead in a huge way.
Other cost cutting measures have been moodle over blackboard, Google Mail/ Cal instead of exchange or groupwise, truecrypt instead of commercial encryption for our notebooks, and virtualbox instead of vmware workstation for our IT student's labs.
I've also used virtualization to save us large amounts of money and at the same time giving us the flexibility to change rapidly instead of our old model of no big changes except for once a year. The biggest cost cut was solaris to linux. Instead of running solaris on sparc hardware and paying 10,000.00 a server. We are running linux on intel hardware and paying 3000.00 a server. With our new virtualization project we are going to reduce that cost even more. I'm installing a blade system that will handle our entire physical server environment + some for the same cost it would take to replace our aging 5 year old servers. Obviously we will still have physical servers, but ESX + linux has reduce the overhead in a huge way.
Other cost cutting measures have been moodle over blackboard, Google Mail/ Cal instead of exchange or groupwise, truecrypt instead of commercial encryption for our notebooks, and virtualbox instead of vmware workstation for our IT student's labs.
MagnusTheBrewer
IN MEMORIAM
- Jun 19, 2004
- 24,122
- 1,594
- 126
You also have to realize, corporations aren't in the business of limiting risk, they're in the business of avoiding it. Open source is an unknown and therefore risky endeavor so no open source for the mother ship. Also, corporate drones aren't payed bonuses for implementing new policies, their paid for meeting the goals of the existing ones.
Nice, this is the kind of projects that would really interest me to see happen. Of course, it has to be done in baby steps.
I've heard that active development in OpenOffice has lagged. While not stalled completely, it seems close. This is the nature of a lot of open source software, even the wildly popular projects. Few companies wish to invest (time or money) in something susceptible to the vagaries of a developer's attention span.
And to those that say "if there's a problem, just dig into the source": this statement fails on so many levels I don't even know where to begin.
And to be clear, I really dig a lot of OSS.
And to those that say "if there's a problem, just dig into the source": this statement fails on so many levels I don't even know where to begin.
And to be clear, I really dig a lot of OSS.
I thought the business model of linux and other open source s/w was to give away the s/w and charge for support. So, is paid support from ibm, redhat and ubuntu (to name a few) worthless? I don't know since I've never used it. It would be interesting to hear from someone who's not speculating.
You have malformed ideas of what happens in the real corporate world. Not sure where you get this "The man is keeping me down" attitude that you espouse all over.
Truth is, a great number of companies, large and small, adopt open-source in some capacity. Open-source doesn't just mean Linux you understand. There are a great number of open-source projects out there that companies adopt, from JBoss to Eclipse.
Like everything else in the world, the realities of open-source have settled with the realities of commerce. It's used where it's beneficial and that's about it. The idea that it has to be all open-source or all Microsoft/Sun/whomever is just ideological nonsense.
Where I work we have a place to submit questions that get answered by a leadership person and posted. One of the recent ones was asking why not switch to OpenOffice. It mainly was due to compatibility and learning curve.
Personally I'm glad because I hate it.
I am impressed and love what open source people do, but I just can't deal with OpenOffice, especially for Excel which we use heavily of course.
But for other programs, I wouldn't mind a OO alternative. Although I can't really think of any that could replace anything else I use at work because most of the major programs I use aren't going to have a OO alternative. I do think we use Dia though because of the cost of Visio licenses. As a software guy, I need a diff tool. I'm sure there's OO ones out there but from the few free ones I've used, Beyond Compare is far superior.
But yeah, commercial products cost, a lot. We use MATLAB a lot here and I've heard some numbers thrown around of what these things cost and it is insane. And there are so many components to it like Simulink, RTW, etc. Plus a 25% yearly support/maintenance fee on every license. :Q
Personally I'm glad because I hate it.
I am impressed and love what open source people do, but I just can't deal with OpenOffice, especially for Excel which we use heavily of course.But for other programs, I wouldn't mind a OO alternative. Although I can't really think of any that could replace anything else I use at work because most of the major programs I use aren't going to have a OO alternative. I do think we use Dia though because of the cost of Visio licenses. As a software guy, I need a diff tool. I'm sure there's OO ones out there but from the few free ones I've used, Beyond Compare is far superior.
But yeah, commercial products cost, a lot. We use MATLAB a lot here and I've heard some numbers thrown around of what these things cost and it is insane. And there are so many components to it like Simulink, RTW, etc. Plus a 25% yearly support/maintenance fee on every license. :Q
Like ms-ofc 2003 vs. ms-ofc 2007?Originally posted by: duragezic
Where I work we have a place to submit questions that get answered by a leadership person and posted. One of the recent ones was asking why not switch to OpenOffice. It mainly was due to compatibility and learning curve.
Personally I'm glad because I hate it.
But for other programs, I wouldn't mind a OO alternative. Although I can't really think of any that could replace anything else I use at work because most of the major programs I use aren't going to have a OO alternative. I do think we use Dia though because of the cost of Visio licenses. As a software guy, I need a diff tool. I'm sure there's OO ones out there but from the few free ones I've used, Beyond Compare is far superior.
But yeah, commercial products cost, a lot. We use MATLAB a lot here and I've heard some numbers thrown around of what these things cost and it is insane. And there are so many components to it like Simulink, RTW, etc. Plus a 25% yearly support/maintenance fee on every license. :Q
Not for us. The reason we do not use open source, we have no one to call when it goes down. Note, we do have open source. However, we only have open source in places where it is supported. If the software is supported on an open source OS, then we are free to use it. If the vendor will not support it, that is it. End of discussion, open source just lost. BTW, we are a health care facility. We once had an update bring down a critical system that affected patient care. Believe me, when the CEO, the CIO, and the head of nursing stop by your desk to ask when it will be back up, you want to be able to say "X from support is looking at this now, we are at Y stage of testing the fix, and we hope to be back up in Z" I cannot imagine saying "I asked on this forum, and I hope to get a response back today." Some businesses may be different, but believe me, when the stuff stops working, and something important is on the line, you need someone you can yell at, and his superior if he does not move fast enough.
It's important to note that "open source" is a very broad term that covers a wide variety of business models. There's software that is completely free, including the support (or lack thereof). There are some incredibly awesome programs that fall under this category with great followings and very active forums. But even in the best case scenarios, there is still a risk of encountering an obscure bug that no one else has found, and in those cases, you have no idea when you're going to get a fix. There's no SLA, since you're not paying for a service. No matter how mission critical your problem is, no one is realistically on the hook for getting you a solution - ever. That's not saying a fix won't eventually come out, but there is no guarantee when it comes to completely free software and support.
Of course, some of you guys are like, "just look at the source code, stupid!" Honestly, maybe I'm just not as great of a programmer as some of you, but the code for a lot of very complex programs is...ummm...complex. As in, unless you were actively participating in developing that software, you probably have no f-ing clue what you're doing. You may introduce more problems than you "fix." It takes weeks/months to completely ramp up new developers for some of our more complicated projects, and that's usually with guidance from the original programmers. Now, during a production emergency, you're hoping that one of your software devs (not sysadmins, since there's a difference) can take the time to go digging through the source code, hoping/praying that there's SOME comments/documentation...
Some open source companies, as someone mentioned above, give away the software, but sell support. Red Hat (full disclosure: my company is a subsidiary of Red Hat) works on a variation of that model. Yes, they have software that is free, such as Fedora or JBoss. Those "community" versions have 0 support, although fixes and new features are constantly being released. Rapid development occurs in the community/free releases, but no guarantee is made on stability or even the functionality. Red Hat then takes promising releases and after some GA and CR iterations, produce an enterprise version of their software. People pay for the enterprise license, and with that, finally, comes support. This mitigates some of the risk that corporations traditionally see with open source software. You'll hopefully get a lot less "what nightly build are we patching from" or "I hope this version of x works with that version of y."
Overall, I think proprietary and open source software purchasing decisions should be made based on needs, not on ideology.
Of course, some of you guys are like, "just look at the source code, stupid!" Honestly, maybe I'm just not as great of a programmer as some of you, but the code for a lot of very complex programs is...ummm...complex. As in, unless you were actively participating in developing that software, you probably have no f-ing clue what you're doing. You may introduce more problems than you "fix." It takes weeks/months to completely ramp up new developers for some of our more complicated projects, and that's usually with guidance from the original programmers. Now, during a production emergency, you're hoping that one of your software devs (not sysadmins, since there's a difference) can take the time to go digging through the source code, hoping/praying that there's SOME comments/documentation...
Some open source companies, as someone mentioned above, give away the software, but sell support. Red Hat (full disclosure: my company is a subsidiary of Red Hat) works on a variation of that model. Yes, they have software that is free, such as Fedora or JBoss. Those "community" versions have 0 support, although fixes and new features are constantly being released. Rapid development occurs in the community/free releases, but no guarantee is made on stability or even the functionality. Red Hat then takes promising releases and after some GA and CR iterations, produce an enterprise version of their software. People pay for the enterprise license, and with that, finally, comes support. This mitigates some of the risk that corporations traditionally see with open source software. You'll hopefully get a lot less "what nightly build are we patching from" or "I hope this version of x works with that version of y."
Overall, I think proprietary and open source software purchasing decisions should be made based on needs, not on ideology.
on top of lack of support, one other reason i've been given is the fact that it is open source. meaning, that anyone and everyone has access to the source code for what ever app is being used, hence scripts, bugs and such can be used and attacked much easier.
i personally don't know how much of it is true since i really only dabble in open source.
i personally don't know how much of it is true since i really only dabble in open source.
James Bond
Diamond Member
- Jan 21, 2005
- 6,023
- 0
- 0
My parent company is in the middle of switching almost all servers to Unix.
We are beginning by replacing all on-site Windows DC's (30 sites) - the amount they will save will be huge. Licensing, software, hardware requirements, oh my.
Edit: This will eventually roll over to users. Currently planning on replacing Office with Open Office and we are discussions of replacing Exchange with Google Mail (using docs, calendars, everything) - not so sure how the last part will go.
We are beginning by replacing all on-site Windows DC's (30 sites) - the amount they will save will be huge. Licensing, software, hardware requirements, oh my.
Edit: This will eventually roll over to users. Currently planning on replacing Office with Open Office and we are discussions of replacing Exchange with Google Mail (using docs, calendars, everything) - not so sure how the last part will go.
SphinxnihpS
Diamond Member
- Feb 17, 2005
- 8,368
- 25
- 91
Like there is support from MS? There is if you pay for it, but even then, those people are stupid and waste your time.
Because the guy who decides needs someone who he can blame if something goes wrong. "Blame the vendor" is a very easy strategy for IT managers.
There is even an old 80's saying that is still very true
"Nobody ever got fired for buying IBM"
just replace IBM by Microsoft, Oracle, Solaris, ...
There is even an old 80's saying that is still very true
"Nobody ever got fired for buying IBM"
just replace IBM by Microsoft, Oracle, Solaris, ...
Haha I guess I'm hypocritical there because I talk about the learning curve with OO but can't get by people who don't pick up Office 2007 pretty easily. I really like what they did with 2007 and had no problem picking up almost everything I used commonly. One of the reasons we don't have 2007 at work is this.Originally posted by: seemingly random
Like ms-ofc 2003 vs. ms-ofc 2007?Originally posted by: duragezic
Where I work we have a place to submit questions that get answered by a leadership person and posted. One of the recent ones was asking why not switch to OpenOffice. It mainly was due to compatibility and learning curve.
Personally I'm glad because I hate it.
But for other programs, I wouldn't mind a OO alternative. Although I can't really think of any that could replace anything else I use at work because most of the major programs I use aren't going to have a OO alternative. I do think we use Dia though because of the cost of Visio licenses. As a software guy, I need a diff tool. I'm sure there's OO ones out there but from the few free ones I've used, Beyond Compare is far superior.
But yeah, commercial products cost, a lot. We use MATLAB a lot here and I've heard some numbers thrown around of what these things cost and it is insane. And there are so many components to it like Simulink, RTW, etc. Plus a 25% yearly support/maintenance fee on every license. :Q
what OS are those DC's being replaced? that just seems like wasted money to me unless you have a plan for all of those licenses aside from directory services. i imagine you're using the same hardware the windows boxes were in? and what about UNIX support, you have trained staff for that already in place?
wow i hope this was thought out very carefully. good luck with that.
- Oct 28, 1999
- 62,484
- 8,345
- 126
If I remember correctly, the OP works in healthcare. For us it isn't support of the direct app that is installed, it's also validation and support from other vendor apps that reside on that machine.
The healthcare vendors we use have very strict lists of programs that they validate with their client applications to peacefully coexist on the same workstation. These lists are very limited to very widely accepted applications and will almost never include open source apps.
If their application blows up and they see that we have unvalidated applications on that machine they'll deny support without charging us an enourmous fee.
The healthcare vendors we use have very strict lists of programs that they validate with their client applications to peacefully coexist on the same workstation. These lists are very limited to very widely accepted applications and will almost never include open source apps.
If their application blows up and they see that we have unvalidated applications on that machine they'll deny support without charging us an enourmous fee.
You can get people who are very good at Linux and administration, but you will have to browse around and it takes a while. Even if you get people who are great at dealing with these things, most people want a set-it-and-forget-it mode. You'll basically invest in people dedicated to keep your software running and working. The goal of IT is to keep things smooth running but to upgrade infrastructure and improve it to match a company's growth. You're going to ask for a much larger IT department in order to keep things smoothly running.
It's just like why a lot of us nerds buy prebuilt laptops and desktops for our parents and relatives. There are times when we give them hand-me-downs, but usually when people ask for something new, I just recommend a prebuilt product. Building a computer is just small scale, but even then you can run into integration challenges. Now try it for a whole company. You're bound to have more than a few integration challenges.
James Bond
Diamond Member
- Jan 21, 2005
- 6,023
- 0
- 0
It's a non-profit company that receives grants from the state, so most machines (servers and user-PC's) are very outdated. Either way, it is time for an upgrade, so it isn't really like they're throwing away decent software.
Most servers are currently running Win2k with 100GB HDD's - only 60GB of that being used for user data; paired with 512MB Ram and P4's, yuck.
I'm on the networking side, so I'm not involved with the planning for this. Put simply, I really disagree with the direction they are going due to support alone - it will be a nightmare. And no - the Sys team does not have adequate training. Oh well - not my problem... Yet.
From a finance perspective nickel and diming on capex won't save heads nor will it make much of a difference on the financial statements. Performance is generally measured through EBITDA, which excludes depreciation of software/hardware. These days it's headcount that is the primary cost savings/driver.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 20K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter