I can also set it up so that, instead of using a swap file, it uses linux's disk swap partition. And that, I suspect, is there the performance comes from.
The 'swap file' is a truly horrible idea: a fragmentable file on top of a journaled filesystem, vs. a slice of the disk which can be used as directly addressable swap. HUGE difference in performance there, and it's pathetic that MS still hasn't fixed this glaring inadequacy.
I would rather the government hold my money (at "0% interest", which isn't quite a possibility but I'll bite) at a higher rate - and give it back to me after they've made use of it - than for them to simply take it, as happens with so many people.
If the 'tax rate' was raised to something like 50% for everyone (or dropped, as the case may be) and then the government had unlimited use of said money for a year, returning the money to its owner, it would only do good things for the economy and the government, I think. You'd have wild bursts of economic stimulus, you'd have a more responsible government, and the government would have a genuine interest in seeing people succeed instead of keeping them on the dole.
Windows 7 with VS2010 won't even run on that hardware; why would you expect it to work on a virtual environment? That's somewhat unreasonable.
(Though, I will say from personal experience, it's amazing how much better a Windows VM will run on the same hardware it was on prior, but virtualized and with slightly less RAM due to system overhead. I did this once some time ago - w2k3 w/ VS2008, 400M RAM with the rest for the CentOS host. Disk access was, sadly and amazingly, faster and swapping wasn't half as painful as it was in "just" Windows.)
If he does get rid of Windows, then there's no point in using images at all. It's a waste of storage space, and a headache to maintain. Just set up a locally cached storage repository (which you then maintain/keep up to date by manually clearing packages) and install from that using a package list. Use configuration management (something like puppet).
Of course, for 20 systems, that's overkill. For the time that a Linux install takes, simply having a local mirror would likely be Good Enough. If there are no issues with 3rd party apps or esoteric version requirements, that's the ticket.
At any rate, virtualization at the workstation level to abstract the primary utility is the Wrong Approach.
One thing I've learned is that simplicity is often better than complexity. KISS. This plan doesn't: while it might save some time on deployment of a new system, it's needlessly complex and has twice as much maintenance involved. Also, it adds additional headaches due to Windows licensing (unless they're going to sysprep the machines).
There are a lot of little "gotchas" which someone not up on such things might overlook. To be informed on such things, you've either got to do copious amounts of research, test a sizable setup yourself, and/or have experience with a similar deployment. All of these things are reasons to keep each part of a system as simple as reasonably possible (without compromising its functionality).
Lack of experience? Or a lack of give-a-damn or understanding? I've seen 30+ year old "computer engineers" with a decade of experience as system administrators pull this nonsense. They've got experience, they've just got the wrong mindset.
They're like the bad mechanic who replaces parts until finding the problem, or until the problem stops manifesting itself - regardless of whether the problem has actually been fixed.
Virtualization is not a cure-all (and your approach is wrong, to boot).
What you're looking to do is use the latest, greatest technology for profit(!!!). You're going about it wrong. There are plenty of other, better technologies to accomplish the same basic thing. Proper system imagining/installation via something like an installation server.
When you've got 20 workstations, you're at that cusp of continuing on the path you're on (and hopefully, resorting to a method of consistent repeatability) or deciding on a different approach - thin clients, perhaps. Or maybe virtualization is the right approach - but I can guarantee that there's likely no good reason to virtualize Windows on top of each of the 20 workstations that couldn't be solved with better design.
Honestly, if you're one of multiple IT in a place with only 20 workstations, you're seriously over-staffed. Someone - if not you, someone else - is going to figure this out, and figure out a way to make themselves important and you redundant. Even with moderate consistency and controls, a single competent Administrator should be able to take care of 5 times as many workstations and a handful of servers without too much sweat.
But it was a terrible pain to install for a young amateur compared to just popping a LiveCD today. Have fun partitioning your HD with raw fdisk (cfdisk if lucky) and setting up XFree86 by hand to see any graphics. Try setting up non-PNP ISA devices with screwy drivers -- often you had to go hardware swapping for something specific, like a $10 Crystal Sound card. Try rebuilding the Kernel with an ALSA patch to get that to run. Try not using a packaging system for anything -- RPM was terrible at the time, you were better off just compiling things.
That was closer to 12-14 years ago than it was 10 years ago. Ten years ago there was Debian, Progeny, and Stormix. There were G400 cards, and ALSA didn't exist yet.
If the post-secondary education market is segmented ideally, CS graduates should be have the knowledge and conceptual tools to run rings around technical graduates in terms of understanding and designing principles and solutions in a broader corporate or societal context.
Sadly, my experience has been in strong contrast to this (within the domain of systems administration).
Yes, there are university grads who know their stuff and can hack it, too. But from what I've seen, they're got absolutely no idea what they're doing around a system, and most of their administration attempts are utter crap laziness (at best).
In short, they design their systems like a conceptual mathematics student or a computer science type might: no forethought into the future, at all. They're not capable of a true "engineering" task, because everything they touch is of the "it works perfectly, just don't touch it" variety.
In short, the work is good, btu the conceptual understanding of what's necessary over time is not. They've got too much theory floating around in their head and, while they can grasp technologies well enough, proper design isn't a concern if the "theory" is fun enough.
The Black Hand wanted to provoke war between the countries, or at least encourage turmoil. They saw the turmoil as a good way to shove their country towards the Bolshevik Utopia of Marx and Stalin.
The good guys lost. No, the Bad Guys didn't reach their goal, per se, but they did give birth to continually expanding Soviet politics for the better part of several generations.
And now Soviet-era politics are on a rebound worldwide. We're still in the same essential war: it's not a nationalist or international war, so much as it's a cultural one of competing agendas and ideals. On one side you've got the instigators and Marxian revolutionaries; on the other, you've got the Old Guard and/or capitalism/republic/monarchy/fiefdom.
Eh, back circa ~2000 there were the usual SMB shares used by most students. You'd find the occasional decent collection of software, porn, music, and movies this way.
The entire campus (small private college) had a T1 to share, so people did use these shares liberally. Even then, things could get a little congested (I seem to recall that our dorm was on two 48-port 10BT hubs). I doubt it'd have been half as popular if we had bandwidth to spare with things like Pandora and Hulu.
However, there was also the "OOB" file sharing: you had to know the server owner(s) to gain access, which usually meant you had to be a friend or a friend of a friend who had a sizable amount of data to contribute. Eventually the system(s) got moved to the campus server room where the better switching equipment was, on account of one of the server owners being a senior and a sysadmin there.
I'm not so sure I'm as nostalgic about it as you are, but then I was sharing files over SMB on my cable provider two years before that, back at home, at close to 10-base-T speeds. And these days, the effort to find stuff online is comparable to what it was back then on the private networks - assuming you knew the right people - but there's a lot more of it, and much of it is being offered by its owners, free for the taking or for a nominal cost (porn, movies, TV shows, etc.). It's not worth the effort anymore.
It's a bit of a catch-22. Most of those bands would not exist without the immediate dispersal methods of file sharing: people aren't likely to plop any sum of money down on an 'unknown'.
Are they 'entitled' to profit from their ventures when people like them? No, they're not; that's not how it works in this world. Should they be compensated by those who like their work? Of course - if they want to continue to see the fruits of those people's labors (assuming those people are not content to work for free).
It's a trade-off of sorts. You can't have both bounty and high cost in a medium which is, essentially, free for the taking. Human nature doesn't allow for it (and I'd argue, laws to the contrary are immoral).
Yep. When I was in high school just a little over a decade ago (and my siblings a couple years after that), the MO for gaining explicit system privilege was as follows:
* observe the "education major" type in his/her password from the front row. * observe "teacher's" password, sticky-noted to the display, keyboard, or desk in plain view. * watch the teacher hen-peck the password in * failing those, find a friend who did the same and didn't have the cojones to use said password. * failing a password, chances are there's at least 2-3 systems on the school network which still have the default administrative passwords (or no such passwords) set. If you can gain a "system" level account in that fashion (trivial), chances are you'll be able to hop along from one of the user accounts to another system's administrative account using a teacher's password.
More often than not, seems a lot of the teachers' accounts were Windows Administrator accounts (domain or local), and they'd have "all" access with a default view (or something like that) in most of their applications. Educator's software is not typically something given much "security conscious" attention on the design and default setup... that'd make it difficult, and these are educators, after all.
Being a student in (public) school and "taking advantage of the system" was pretty easy. It only took a couple basic assumptions:
* If you're of average intelligence or better, you're smarter than most of your teachers. These are least-common denominator/public educators we're talking about. * Despite the above (and their years in humanities, child psychology, and differential comparison courses), the general attitude of teachers is "kids are stupid". Exploit this. * Teachers assume kids need "direction" not "discipline" when misbehaving, resulting in an escalation of disciplinary issues. This makes all but the most serious things "overlookable" - because if they didn't, they'd spend all of their time "directing" kids to "do the right thing".
If the health bill were an actual health bill, it could have been shorter. It wasn't so much a reform bill, in the traditional sense of the phrase "political reform" as it is reform in the "modeling clay" sense.
It puts more prohibitions in place for doctors and their clients while at the same time subsidizing the (out of control) insurance companies. Hardly a good shake for you and me.
Honestly, we should root for Adobe on this one. Closed platforms are not our friends. That doesn't mean I like Adobe, but Apple is certainly wrong in this case. Have that put in (legal) writing and it opens the door for Apple's platform becoming more open.
At the same time, someone needs to bring a class action suit against Adobe for Flash. Virulent updates? Check. Wanton security issues with no apparent attempts to fix them (architecturally)? Check. Poor reliability? Check. Exclusionary practices? Check.
If I hire company A to do work for me, and I really like John Doe's work (who works for company A) there should be no reason why I shouldn't be able to hire him if he were to leave company A.
On that part we agree. However, you seem to be unable to see the discrepancy: why can't John Doe pull from the people he's done work for and use them as a client base of his own? He may have made those contacts while working for company A, but they're still his contacts (just as he is a contact of your company).
Not allowing an individual to do what a company so frequently does is unscrupulous. Not only that, but it's probably less ethical: as an employer, you're getting not only him but his contacts, and "compensating" him for it. If he (an individual) were to leave, he'd not be acting as an "income intermediary" by pulling off clients from previous employers: he'd be using his own equity in those clients for himself directly.
* Up-front costs might be less, but long-term costs are significantly higher than both having the work locally-sourced (or to the Midwest) or done in-house. * The costs of bad code cascade. It's not as simple as business requirements being met: you've got to contend with significant bugs, memory leaks, and general system performance. Those are all issues which require significant support, much of which can't be provided by someone in India. * If you throw a job at someone in India, you've got no way to enforce quality. Sure, you might have a requirement in the contract saying "code must be commented, meet XYZ standards, and do CYA" but there's no guarantee they're not using ass-backwards logic to perform those tasks, and the likelihood that the code is actually commented to a degree that is useful is slim-to-nill. * Again, back to the documentation: if you want more than just a throw-away program/system, good documentation is existential. The amount of time to reverse-engineer something is exponential to the time required to do it in the first place.
What existing package manager are you modeling your system after, if any?
Do you have a feature list which you will be implementing (or already have implemented)? For instance: pinning, multiple versions of the same package, multiple "satisfies" vectors, etc.
The lack of proper package management in Windows (for all applications, but particularly Microsoft software) has been a huge beef of mine for the better part of a decade. Glad to see Windows is finally getting it! It'll bring Windows' utility as a server OS up quite a bit (probably up to or above that of something like FreeBSD or gentoo).
I just dont think MS is overly concerned that it will compete with their software ecosystem at this point, and is more convinced that government regulators are the bigger threat.
Yep, exactly. As it pertains to Sharepoint, the more something sucks, the more it seems to get broad adoption early on.
Quite smart of Microsoft to make Sharepoint a "killer app" at that - all-encompassing and utterly impossible to replace in whole.
Slashdot Mirror
I can also set it up so that, instead of using a swap file, it uses linux's disk swap partition. And that, I suspect, is there the performance comes from.
The 'swap file' is a truly horrible idea: a fragmentable file on top of a journaled filesystem, vs. a slice of the disk which can be used as directly addressable swap. HUGE difference in performance there, and it's pathetic that MS still hasn't fixed this glaring inadequacy.
I would rather the government hold my money (at "0% interest", which isn't quite a possibility but I'll bite) at a higher rate - and give it back to me after they've made use of it - than for them to simply take it, as happens with so many people.
If the 'tax rate' was raised to something like 50% for everyone (or dropped, as the case may be) and then the government had unlimited use of said money for a year, returning the money to its owner, it would only do good things for the economy and the government, I think. You'd have wild bursts of economic stimulus, you'd have a more responsible government, and the government would have a genuine interest in seeing people succeed instead of keeping them on the dole.
Windows 7 with VS2010 won't even run on that hardware; why would you expect it to work on a virtual environment? That's somewhat unreasonable.
(Though, I will say from personal experience, it's amazing how much better a Windows VM will run on the same hardware it was on prior, but virtualized and with slightly less RAM due to system overhead. I did this once some time ago - w2k3 w/ VS2008, 400M RAM with the rest for the CentOS host. Disk access was, sadly and amazingly, faster and swapping wasn't half as painful as it was in "just" Windows.)
Xen would be the way to do it, if you had servers. Running the display on the same system as the Xen system is, last I checked, not yet possible.
If he does get rid of Windows, then there's no point in using images at all. It's a waste of storage space, and a headache to maintain. Just set up a locally cached storage repository (which you then maintain/keep up to date by manually clearing packages) and install from that using a package list. Use configuration management (something like puppet).
Of course, for 20 systems, that's overkill. For the time that a Linux install takes, simply having a local mirror would likely be Good Enough. If there are no issues with 3rd party apps or esoteric version requirements, that's the ticket.
At any rate, virtualization at the workstation level to abstract the primary utility is the Wrong Approach.
One thing I've learned is that simplicity is often better than complexity. KISS. This plan doesn't: while it might save some time on deployment of a new system, it's needlessly complex and has twice as much maintenance involved. Also, it adds additional headaches due to Windows licensing (unless they're going to sysprep the machines).
There are a lot of little "gotchas" which someone not up on such things might overlook. To be informed on such things, you've either got to do copious amounts of research, test a sizable setup yourself, and/or have experience with a similar deployment. All of these things are reasons to keep each part of a system as simple as reasonably possible (without compromising its functionality).
Lack of experience? Or a lack of give-a-damn or understanding? I've seen 30+ year old "computer engineers" with a decade of experience as system administrators pull this nonsense. They've got experience, they've just got the wrong mindset.
They're like the bad mechanic who replaces parts until finding the problem, or until the problem stops manifesting itself - regardless of whether the problem has actually been fixed.
Virtualization is not a cure-all (and your approach is wrong, to boot).
What you're looking to do is use the latest, greatest technology for profit(!!!). You're going about it wrong. There are plenty of other, better technologies to accomplish the same basic thing. Proper system imagining/installation via something like an installation server.
When you've got 20 workstations, you're at that cusp of continuing on the path you're on (and hopefully, resorting to a method of consistent repeatability) or deciding on a different approach - thin clients, perhaps. Or maybe virtualization is the right approach - but I can guarantee that there's likely no good reason to virtualize Windows on top of each of the 20 workstations that couldn't be solved with better design.
Honestly, if you're one of multiple IT in a place with only 20 workstations, you're seriously over-staffed. Someone - if not you, someone else - is going to figure this out, and figure out a way to make themselves important and you redundant. Even with moderate consistency and controls, a single competent Administrator should be able to take care of 5 times as many workstations and a handful of servers without too much sweat.
But it was a terrible pain to install for a young amateur compared to just popping a LiveCD today. Have fun partitioning your HD with raw fdisk (cfdisk if lucky) and setting up XFree86 by hand to see any graphics. Try setting up non-PNP ISA devices with screwy drivers -- often you had to go hardware swapping for something specific, like a $10 Crystal Sound card. Try rebuilding the Kernel with an ALSA patch to get that to run. Try not using a packaging system for anything -- RPM was terrible at the time, you were better off just compiling things.
That was closer to 12-14 years ago than it was 10 years ago. Ten years ago there was Debian, Progeny, and Stormix. There were G400 cards, and ALSA didn't exist yet.
I had the same outlook about the Linux kernel, but the opposite: I thought it was an understandable fiend, and started to delve in.
This was when 2.6 was in development, and 2.4 was the main branch. And then 2.6 came out, and all was lost...
If the post-secondary education market is segmented ideally, CS graduates should be have the knowledge and conceptual tools to run rings around technical graduates in terms of understanding and designing principles and solutions in a broader corporate or societal context.
Sadly, my experience has been in strong contrast to this (within the domain of systems administration).
Yes, there are university grads who know their stuff and can hack it, too. But from what I've seen, they're got absolutely no idea what they're doing around a system, and most of their administration attempts are utter crap laziness (at best).
In short, they design their systems like a conceptual mathematics student or a computer science type might: no forethought into the future, at all. They're not capable of a true "engineering" task, because everything they touch is of the "it works perfectly, just don't touch it" variety.
In short, the work is good, btu the conceptual understanding of what's necessary over time is not. They've got too much theory floating around in their head and, while they can grasp technologies well enough, proper design isn't a concern if the "theory" is fun enough.
When America holds true to its founding principles, every endeavor pursued by the US has been triumphant.
This healthcare bullshit flies right in the face of said principles (as does pretty much everything Obama and the current congress has done).
WWI was about "good vs. bad".
The Black Hand wanted to provoke war between the countries, or at least encourage turmoil. They saw the turmoil as a good way to shove their country towards the Bolshevik Utopia of Marx and Stalin.
The good guys lost. No, the Bad Guys didn't reach their goal, per se, but they did give birth to continually expanding Soviet politics for the better part of several generations.
And now Soviet-era politics are on a rebound worldwide. We're still in the same essential war: it's not a nationalist or international war, so much as it's a cultural one of competing agendas and ideals. On one side you've got the instigators and Marxian revolutionaries; on the other, you've got the Old Guard and/or capitalism/republic/monarchy/fiefdom.
Eh, back circa ~2000 there were the usual SMB shares used by most students. You'd find the occasional decent collection of software, porn, music, and movies this way.
The entire campus (small private college) had a T1 to share, so people did use these shares liberally. Even then, things could get a little congested (I seem to recall that our dorm was on two 48-port 10BT hubs). I doubt it'd have been half as popular if we had bandwidth to spare with things like Pandora and Hulu.
However, there was also the "OOB" file sharing: you had to know the server owner(s) to gain access, which usually meant you had to be a friend or a friend of a friend who had a sizable amount of data to contribute. Eventually the system(s) got moved to the campus server room where the better switching equipment was, on account of one of the server owners being a senior and a sysadmin there.
I'm not so sure I'm as nostalgic about it as you are, but then I was sharing files over SMB on my cable provider two years before that, back at home, at close to 10-base-T speeds. And these days, the effort to find stuff online is comparable to what it was back then on the private networks - assuming you knew the right people - but there's a lot more of it, and much of it is being offered by its owners, free for the taking or for a nominal cost (porn, movies, TV shows, etc.). It's not worth the effort anymore.
It's a bit of a catch-22. Most of those bands would not exist without the immediate dispersal methods of file sharing: people aren't likely to plop any sum of money down on an 'unknown'.
Are they 'entitled' to profit from their ventures when people like them? No, they're not; that's not how it works in this world. Should they be compensated by those who like their work? Of course - if they want to continue to see the fruits of those people's labors (assuming those people are not content to work for free).
It's a trade-off of sorts. You can't have both bounty and high cost in a medium which is, essentially, free for the taking. Human nature doesn't allow for it (and I'd argue, laws to the contrary are immoral).
Yep. When I was in high school just a little over a decade ago (and my siblings a couple years after that), the MO for gaining explicit system privilege was as follows:
* observe the "education major" type in his/her password from the front row.
* observe "teacher's" password, sticky-noted to the display, keyboard, or desk in plain view.
* watch the teacher hen-peck the password in
* failing those, find a friend who did the same and didn't have the cojones to use said password.
* failing a password, chances are there's at least 2-3 systems on the school network which still have the default administrative passwords (or no such passwords) set. If you can gain a "system" level account in that fashion (trivial), chances are you'll be able to hop along from one of the user accounts to another system's administrative account using a teacher's password.
More often than not, seems a lot of the teachers' accounts were Windows Administrator accounts (domain or local), and they'd have "all" access with a default view (or something like that) in most of their applications. Educator's software is not typically something given much "security conscious" attention on the design and default setup... that'd make it difficult, and these are educators, after all.
Being a student in (public) school and "taking advantage of the system" was pretty easy. It only took a couple basic assumptions:
* If you're of average intelligence or better, you're smarter than most of your teachers. These are least-common denominator/public educators we're talking about.
* Despite the above (and their years in humanities, child psychology, and differential comparison courses), the general attitude of teachers is "kids are stupid". Exploit this.
* Teachers assume kids need "direction" not "discipline" when misbehaving, resulting in an escalation of disciplinary issues. This makes all but the most serious things "overlookable" - because if they didn't, they'd spend all of their time "directing" kids to "do the right thing".
Biggest competitor to Apache out there is Microsoft, as I see it. They're pushing Sharepoint hard, and Apache is the foundation of their competition.
A lot of government entities (US and otherwise) could also benefit, though if it's been detected I'd guess it was somewhat more amateur than that.
As someone who uses Ubuntu on his laptops and workstations, let me just say:
They should've been using a real OS.
I mean, seriously. Ubuntu on a server? Yeah, it's great and all, but it doesn't get a whole lot of scrutiny for security.
If the health bill were an actual health bill, it could have been shorter. It wasn't so much a reform bill, in the traditional sense of the phrase "political reform" as it is reform in the "modeling clay" sense.
It puts more prohibitions in place for doctors and their clients while at the same time subsidizing the (out of control) insurance companies. Hardly a good shake for you and me.
Honestly, we should root for Adobe on this one. Closed platforms are not our friends. That doesn't mean I like Adobe, but Apple is certainly wrong in this case. Have that put in (legal) writing and it opens the door for Apple's platform becoming more open.
At the same time, someone needs to bring a class action suit against Adobe for Flash. Virulent updates? Check. Wanton security issues with no apparent attempts to fix them (architecturally)? Check. Poor reliability? Check. Exclusionary practices? Check.
If I hire company A to do work for me, and I really like John Doe's work (who works for company A) there should be no reason why I shouldn't be able to hire him if he were to leave company A.
On that part we agree. However, you seem to be unable to see the discrepancy: why can't John Doe pull from the people he's done work for and use them as a client base of his own? He may have made those contacts while working for company A, but they're still his contacts (just as he is a contact of your company).
Not allowing an individual to do what a company so frequently does is unscrupulous. Not only that, but it's probably less ethical: as an employer, you're getting not only him but his contacts, and "compensating" him for it. If he (an individual) were to leave, he'd not be acting as an "income intermediary" by pulling off clients from previous employers: he'd be using his own equity in those clients for himself directly.
A half dozen reasons, or more:
* Up-front costs might be less, but long-term costs are significantly higher than both having the work locally-sourced (or to the Midwest) or done in-house.
* The costs of bad code cascade. It's not as simple as business requirements being met: you've got to contend with significant bugs, memory leaks, and general system performance. Those are all issues which require significant support, much of which can't be provided by someone in India.
* If you throw a job at someone in India, you've got no way to enforce quality. Sure, you might have a requirement in the contract saying "code must be commented, meet XYZ standards, and do CYA" but there's no guarantee they're not using ass-backwards logic to perform those tasks, and the likelihood that the code is actually commented to a degree that is useful is slim-to-nill.
* Again, back to the documentation: if you want more than just a throw-away program/system, good documentation is existential. The amount of time to reverse-engineer something is exponential to the time required to do it in the first place.
You seem to be speaking as if this project is well beyond the officially-sanctioned-with-Microsoft stages.
"Unofficially", how mature are your efforts to making sane package management in Windows a reality? 6 months to public release? A year? Two years?
What existing package manager are you modeling your system after, if any?
Do you have a feature list which you will be implementing (or already have implemented)? For instance: pinning, multiple versions of the same package, multiple "satisfies" vectors, etc.
The lack of proper package management in Windows (for all applications, but particularly Microsoft software) has been a huge beef of mine for the better part of a decade. Glad to see Windows is finally getting it! It'll bring Windows' utility as a server OS up quite a bit (probably up to or above that of something like FreeBSD or gentoo).
I just dont think MS is overly concerned that it will compete with their software ecosystem at this point, and is more convinced that government regulators are the bigger threat.
Yep, exactly. As it pertains to Sharepoint, the more something sucks, the more it seems to get broad adoption early on.
Quite smart of Microsoft to make Sharepoint a "killer app" at that - all-encompassing and utterly impossible to replace in whole.