Excuse me if I sound stupid but consolidation and virtualization could be two separate processes or system management criteria. Depending on the application, having a SQL server of some type or other could conceivably serve multiple database applications. Consolidation could be as simple as just moving a database application(s) onto a partially utilized SQL server and not require any virtualization at all.
Likewise virtualization might allow for some special applications to run that don't have heavy usage or limited appeal. Like using a backup server as a network monitor for instance.
I have seen tremendous amounts of IT-spam recently in the last 6 months advertising, announcing or proselytizing virtualization software. And yes its going to be the browser wars all over, especially if they give it to us free and expect us to buy either support or "enhanced performance" professional versions. Snake oil all of it.
Yes I have used Virtual PC 2007 and tried VMWare and they are fine for what they are and do. But no matter how you cut it, there is a cost of power that can be associated with every bit and byte forced through a Hypervisor/virtual machine/root kit or OS. The CPU is going to run hotter and harder and probably won't last as long. Yes you might save some money by combining tasks or applications onto fewer servers ONLY if you haven't ALREADY installed them onto their own servers. If the servers are still running, do they really use that much AC power and cooling to justify replacing them with something that will run hotter and cost more? Maybe you should have replaced or shutdown those servers and applications already?
What's it going to cost to buy a high powered server, install an OS on it, install a hypervisor (huh? Since when is that a software application or component?) and then install all the not-so-virtual OS licenses onto the system? Then test it to make sure it doesn't bring down the entire house of cards you've made by putting multiple applications onto one system. You want to try and monetize that and then justify it to management? Try generating/calculating a legitimate ROI for an application that doesn't evidently see much usage. If its significant enough to be more than 1% of your total system bandwidth or gross product, I wouldn't screw with it. If the application is generating more than 1% of your business traffic on the network or Internet, you might want to seriously consider mirroring or clustering instead of using virtualization.
If it's generating less than that, WHY RUN IT AT ALL? Give your customers a copy of the software and show them how to run it locally on their server and then offer a remote service contract to maintain it if you have to.
If you're worried about security threats from it, GET IT RE-WRITTEN to eliminate the threats.
I am amazed that the latest buzzword or fad software becomes an answer to all sorts of questions that should have more intelligent answers than just shoving it into a virtual machine.
If the code is buggy, it stays buggy on the VM.
If its a security threat, that doesn't change much either. How long do you think it will take for the malware miscreants to find holes in MS or anybody else's VM software and take advantage of it?
I think that like Vista for Microsoft, virtualization has become the cash-cow everybody can own. All they have to do is buy some software or hire some programmers for the project and then white-paper/advertise the crap out of it and then bingo the cash starts coming in, no Microsoft required. Now Microsoft wants a piece of the market.
The mechanism all of the virtual-fan-boys use is a survey.
One of the problems I have run into answering surveys is that the preset expectations or mindsets of the testers has a tremendous impact on how well designed the survey is and how effective it is at soliciting meaningful and unbiased data. I have given up answering the damn things because its obvious in most cases that the selections set as possible answer choices in most cases will force or bend the survey to a specific goal.
The flood of babel-speak is tremendous. Grab a bunch of buzzwords jam them together in questions that sound more like pick one from column A and 2 from column B and poof you have a tech-geek survey. Its like picking or writing your own set of benchmark tests to demonstrate your CPU is better than a CPU from somebody else. Likewise I don't bother reading the reports from them any longer for the same reasons.
Combining virtualization and consolidation in the same answer selection block prevents finely defining or allowing the difference between the two. Consolidation doesn't require virtual-anything in all cases. The reason I jumped on this specific phrase is that I've seen this combination cited or offered as an answer in more than one survey in the past and more recently as word-fodder in an article promoting or announcing the (now 2 week old) potential release of a Microsoft Hyper-Visor management tool.
source:- http://community.zdnet.co.uk/
Likewise virtualization might allow for some special applications to run that don't have heavy usage or limited appeal. Like using a backup server as a network monitor for instance.
I have seen tremendous amounts of IT-spam recently in the last 6 months advertising, announcing or proselytizing virtualization software. And yes its going to be the browser wars all over, especially if they give it to us free and expect us to buy either support or "enhanced performance" professional versions. Snake oil all of it.
Yes I have used Virtual PC 2007 and tried VMWare and they are fine for what they are and do. But no matter how you cut it, there is a cost of power that can be associated with every bit and byte forced through a Hypervisor/virtual machine/root kit or OS. The CPU is going to run hotter and harder and probably won't last as long. Yes you might save some money by combining tasks or applications onto fewer servers ONLY if you haven't ALREADY installed them onto their own servers. If the servers are still running, do they really use that much AC power and cooling to justify replacing them with something that will run hotter and cost more? Maybe you should have replaced or shutdown those servers and applications already?
What's it going to cost to buy a high powered server, install an OS on it, install a hypervisor (huh? Since when is that a software application or component?) and then install all the not-so-virtual OS licenses onto the system? Then test it to make sure it doesn't bring down the entire house of cards you've made by putting multiple applications onto one system. You want to try and monetize that and then justify it to management? Try generating/calculating a legitimate ROI for an application that doesn't evidently see much usage. If its significant enough to be more than 1% of your total system bandwidth or gross product, I wouldn't screw with it. If the application is generating more than 1% of your business traffic on the network or Internet, you might want to seriously consider mirroring or clustering instead of using virtualization.
If it's generating less than that, WHY RUN IT AT ALL? Give your customers a copy of the software and show them how to run it locally on their server and then offer a remote service contract to maintain it if you have to.
If you're worried about security threats from it, GET IT RE-WRITTEN to eliminate the threats.
I am amazed that the latest buzzword or fad software becomes an answer to all sorts of questions that should have more intelligent answers than just shoving it into a virtual machine.
If the code is buggy, it stays buggy on the VM.
If its a security threat, that doesn't change much either. How long do you think it will take for the malware miscreants to find holes in MS or anybody else's VM software and take advantage of it?
I think that like Vista for Microsoft, virtualization has become the cash-cow everybody can own. All they have to do is buy some software or hire some programmers for the project and then white-paper/advertise the crap out of it and then bingo the cash starts coming in, no Microsoft required. Now Microsoft wants a piece of the market.
The mechanism all of the virtual-fan-boys use is a survey.
One of the problems I have run into answering surveys is that the preset expectations or mindsets of the testers has a tremendous impact on how well designed the survey is and how effective it is at soliciting meaningful and unbiased data. I have given up answering the damn things because its obvious in most cases that the selections set as possible answer choices in most cases will force or bend the survey to a specific goal.
The flood of babel-speak is tremendous. Grab a bunch of buzzwords jam them together in questions that sound more like pick one from column A and 2 from column B and poof you have a tech-geek survey. Its like picking or writing your own set of benchmark tests to demonstrate your CPU is better than a CPU from somebody else. Likewise I don't bother reading the reports from them any longer for the same reasons.
Combining virtualization and consolidation in the same answer selection block prevents finely defining or allowing the difference between the two. Consolidation doesn't require virtual-anything in all cases. The reason I jumped on this specific phrase is that I've seen this combination cited or offered as an answer in more than one survey in the past and more recently as word-fodder in an article promoting or announcing the (now 2 week old) potential release of a Microsoft Hyper-Visor management tool.
source:- http://community.zdnet.co.uk/
Offers Development Services For