My problem is: why should I restart an entire OS when and stop answering to HTTP requests (for example) only the email server needs to be patched? (granted, I made the mistake of co-hosting them on the same box and choosing a Windows OS).
The point is: why should you _care_ if you have to restart a server ? If your architecture is sound then doing so will have no - or extremely minimal - impact.
And this point is: how come the inability of Windows OS-es to handle security patching without a reboot became a case of "broken architecture - not being able to handle planned outages"
It's got nothing whatsoever to do with Windows. If your architecture cannot maintain its SLAs in light of a planned server restart, then it is broken (or your SLAs are inappropriate). This is true no matter what the OS is.
Individual server uptimes are essentially irrelevant outside of e-dick waving. What matters is service availability. A single server cannot be relied on to deliver high levels of availability, regardless of OS.
If the OS running the architecture does not require a reboot after applying security patches, then I don't need to schedule for downtime... no matter how the architecture might be.
You're missing the point. If your architecture can't handle a planned outage of a single server, then it's even less capable of handling an _unplanned_ outage. Ie: it's broken.
Could it be that you haven't yet heard of the "just restart the service" approach or even hot-patching?
Sure. I just have enough experience to know that individual server uptimes are not what matter.
That's not even close to a "Mac Mini Pro". Only a slow dual-core CPU, 8GB RAM max, no PCIe video card, no spare PCIe slot, slow 2.5" drives (and only a single one without sacrificing the optical drive).
What I want is basically half a Mac Pro - or the equivalent of a Dell Precision T1500 if you want something actually on the market. A single CPU socket, up to 16GB RAM, (upgradable) PCIe video card, two free PCIe slots (x4 and x1), two internal 3.5" drive bays and an optical drive.
And I want a base model - quad-core, 4GB RAM, 500GB HDD - that costs about $1300. If Dell can do it for a grand, Apple can do it for $1300 and still collect a reasonable Apple Tax.
Great. So the 6th 9 hacks it and posts a torrent. The cat is out of the bag.
That's of relatively limited importance if it takes them, say, 6-12 months to do it. The bulk of money is going to be made immediately after release, when popularity is high.
The *real* issue is not whether or not the cat gets out of the bag, it's how long it takes to do so.
Why exactly would you be doing real processing on a virtual machine?
Because VMs are much, much nicer to manage than real hardware, if for no other reason than making the ideal one-service-per-server scenario less wasteful.
If you need real processing power you don't put it on a VM, thats just dumb and wasteful.
The overhead of modern virtualisation software on modern hardware is negligible (a couple of percent, if that, outside of corner-case workloads).
Virtual machines are for testing and silly little one off instances of something that some department 'needed' on a 'server' that gets used by 3 people, twice a year, and thats only until next month when they've forgot about it completely.
VMs are for building an infrastructure that's easier to manage and scale than one built on physical hardware. If you're only using VMs the way you describe, you are Doing It Wrong.
In a lot of organizations it is cheaper to pay for a 24/7/365 MySQL expert from the local community. I know a guy who does just that, signs up businesses, gives them the cell phone number, and is prepared at any time to help them.
What does he do if two of them call at the same time ?
The majority of wells show rising levels of pollutant, and the water tables in many places shows a consistent downward trend because we've been using it for irrigation and industrial processes -- and it's not an entirely closed system.
Where ?
But if we run out of drinking water because our wells and lakes are filled with poisons, all the oil in the world won't save our sorry asses. Every human being that ever existed needed water to survive -- daily. It's only the past few iterations of humanity that have needed it, and I'm certain we'll survive as a species without it, and without having to look at making large sections of land uninhabitable or reducing the population to do so. Water is serious business. Oil is a plaything for scientists and engineers to work out an alternative for. There is no alternative for water.
Nor is there any risk of running out of it. Ever. It covers 80% of the surface of the Earth. It falls out of the sky. You piss out 1-2L every day. It can be made potable by applying energy. We successfully build *cities* in the middle of the freakin' desert without even making _basic_ attempts at water conservation and reclamation.
We are not going to run out of water. Worst case is we need to build lots of pipelines, or relocate to places where precipitation is high.
The choice isn't that simple. LVM comes with its own complications, including a tendency to get volume offsets "wrong" so the file system data doesn't align nicely to RAID stripes. This is not good for performance.
This is only going to happen if your partition is misaligned, in which case whether or not you're using LVM is irrelevant.
The solution is to either a) use whole-disk PVs (preferable), or b) create properly aligned partitions.
Also, LVM has only recently acquired barrier support, and the combination of no barriers + write cache can be quite dangerous if power is lost. Even battery backed cache won't save you if you use a journalling file system (and everybody does these days) because request ordering isn't guaranteed.
Again, nothing to do with LVM, such a failure will affect any system.
The vast majority of "need a backup" situations are end users accidentally deleting things. This means a simple self-service restore capability that keeps, say, 1-2 weeks worth of hourly snapshots, will cover 80%-90% of restorations in a very short period of time with nearly no administrative overhead.
You can't do this with LVM (well, you can try, but performance is going to be dismal and it's going to be a nightmare to configure and manage).
And The One would, of course, control everything, and thus be an archetypical instance of a totalitarian government. In fact wasn't that the goal of fascism?
Yes. Hence why removing laws and regulations to prevent that would be a bad thing.
And that's just fine, since any specialized version you might make is also in public domain.
Except for when it's tied down with hardware dongles, DRM, and phone-home schemes. Oh, and you can't get the source code any more, either.
GPL exists to hack around copyright law.
The GPL exists to ensure source code is available to anyone whom binaries are distributed to. It uses Copyright to achieve that goal.
No Copyright == no GPL == no source code available with distributed binaries. If the "source" part of "open source" is important to you, you need Copyright.
Upfront costs are one off costs, and if software was to start expensive and rapidly reduce to near zero cost as the initial costs were covered that would make a lot more sense. You may only have used 10 hours of support time, but you have support time available to you should you need it. Sure they make a profit, but they still need to have support staff on hand to answer your calls. Just because you aren't calling right now doesn't mean the call centre staff can pack up and go home.
No, the costs of initial and ongoing development are amortised across the expected lifetime of the product, and the "up front" pricing set accordingly.
You may only have used 10 hours of support time, but you have support time available to you should you need it. Sure they make a profit, but they still need to have support staff on hand to answer your calls.
Actually it's the updates I consider myself paying for - which is essentially the same situation as proprietary software, only a (higher) annual expense rather than a once-every-few-years cost.
Just because you aren't calling right now doesn't mean the call centre staff can pack up and go home.
Just because $SOFTWARE has been released doesn't mean all those developers can pack up and go home.
The issue is that proprietary software allows ridiculous profit margins (close to 100% since the software costs nothing to distribute and economies of scale are pretty much linear since the upfront costs remain the same regardless of volume)...
That the upfront costs remain the same does not mean they disappear. The proprietary software still needs to make up for its development cost before it's profitable.
You also assume said proprietary software is not maintained and updated after its initial sale (thus incurring more expense), which is patently false.
Open source vendors are unable to rip their customers off by selling zero cost goods at ridiculous markups because if they did someone else could come along and offer the same code for a cheaper price, instead they must make their money selling services... Services have a constant ongoing cost to actually provide the service, and these costs increase as you provide service to more customers.
We pay RH somewhere in the ballpark of fifty grand *per year* for our RHEL licenses. In the last 5 years, we've used maybe 10 hours of support time. I'm pretty sure they've made a lot of profit off us.
So can anyone name any large close source software companies that have started up rather recently that are billion dollar companies? I can't personally think of any. Can anyone else?
VMware.
I expect several of the computer game companies would also make the cut, though acquisitions and such might make that hard to figure out.
Anyway, if copyright laws didn't exist for software? Well, you'd see companies like Microsoft fall apart and companies like Red Hat thrive.
The GPL is utterly reliant on copyright to function as it does. Without it, the most useful GPLed code would be "proprietarised" in short order. Everything GPL proponents criticise the BSDL for would be applicable to GPLed software.
Because the business model would shift from protecting your source code through litigation to making it available for free since that would be the only way to effectively combat piracy.
No, the business model would shift to more effective piracy deterrents like hardware dongles and better DRM.
There is very little piracy of source code, most of it is binary distributions. Companies would probably lock their source code repos up a little tighter, and segregate them more so fewer (if any) employees had access to everything, but most big vendors will be doing this already. The removal of copyright would hurt open source (particularly GPLed) software far more - because while proprietary software vendors are used to dealing with piracy, OSS vendors always work under the assumption that no other OSS vendor will be able to gain a meaningful software-based competitive advantage.
Many businesses that reach billions of dollars in revenue often rely on government contracts and monopoly protection--patent law being the biggest of these. Without government interference in the economy businesses would probably be less likely to hit "billionaire" status. I don't doubt that there would still be some, just not as many.
More likely there'd be just one or two, at trillionaire status, and everything else would be subsidiaries of them.
In the open source world this is (to some extent) playing out.
Open Source software (at least the GPL or similar, which is typically what people mean) is utterly dependent on the "government monopoly protection" of Copyright. Without that, you don't have "Open Source", you have "Public Domain".
The fact is, the only way you would know that your business is serving people is if it makes a profit. Loss-making enterprises mean that there are better uses for your capital.
You can't do that in Democrat-run states (like Maryland) either, so how are they any better?
I'm not making any commentary on whether the Democrats are better, I'm pointing out that the Republicans *clearly* are not interested in "maximum individual freedom" when gender discrimination is a fundamental and ingrained part of their platform.
You still haven't explained why people getting stuff for free because their friends gave it to them, and people getting stuff for free because their friends copied it for them, are different in terms of "ripping off the entertainment you want so you can have it with no money involved".
Slashdot Mirror
The customers are to blame, why should one stick the teeth into Microsoft's neck?
Or RH, or SuSe, or Sun - they're all equally incapable of providing high availability with single-server dependencies.
Oh, do I?
Yes. Even after I explained it.
My problem is: why should I restart an entire OS when and stop answering to HTTP requests (for example) only the email server needs to be patched? (granted, I made the mistake of co-hosting them on the same box and choosing a Windows OS).
The point is: why should you _care_ if you have to restart a server ? If your architecture is sound then doing so will have no - or extremely minimal - impact.
And this point is: how come the inability of Windows OS-es to handle security patching without a reboot became a case of "broken architecture - not being able to handle planned outages"
It's got nothing whatsoever to do with Windows. If your architecture cannot maintain its SLAs in light of a planned server restart, then it is broken (or your SLAs are inappropriate). This is true no matter what the OS is.
Individual server uptimes are essentially irrelevant outside of e-dick waving. What matters is service availability. A single server cannot be relied on to deliver high levels of availability, regardless of OS.
If the OS running the architecture does not require a reboot after applying security patches, then I don't need to schedule for downtime... no matter how the architecture might be.
You're missing the point. If your architecture can't handle a planned outage of a single server, then it's even less capable of handling an _unplanned_ outage. Ie: it's broken.
Could it be that you haven't yet heard of the "just restart the service" approach or even hot-patching?
Sure. I just have enough experience to know that individual server uptimes are not what matter.
What I want is basically half a Mac Pro - or the equivalent of a Dell Precision T1500 if you want something actually on the market. A single CPU socket, up to 16GB RAM, (upgradable) PCIe video card, two free PCIe slots (x4 and x1), two internal 3.5" drive bays and an optical drive.
And I want a base model - quad-core, 4GB RAM, 500GB HDD - that costs about $1300. If Dell can do it for a grand, Apple can do it for $1300 and still collect a reasonable Apple Tax.
Also, customers are also to blame because applying a security patch requires a reboot.
If a planned reboot disrupts services in a meaningful way, then your architecture is broken. This is true regardless of what OS you're running.
In an effort to make the IE a critical part of Windows, all sorts of components of Windows (like the help system) have been shoehorned into IE.
How is using HTML for documentation "shoehorning" ? A help system is pretty much a textbook example of where hyperlinking is a good idea.
Great. So the 6th 9 hacks it and posts a torrent. The cat is out of the bag.
That's of relatively limited importance if it takes them, say, 6-12 months to do it. The bulk of money is going to be made immediately after release, when popularity is high.
The *real* issue is not whether or not the cat gets out of the bag, it's how long it takes to do so.
Why exactly would you be doing real processing on a virtual machine?
Because VMs are much, much nicer to manage than real hardware, if for no other reason than making the ideal one-service-per-server scenario less wasteful.
If you need real processing power you don't put it on a VM, thats just dumb and wasteful.
The overhead of modern virtualisation software on modern hardware is negligible (a couple of percent, if that, outside of corner-case workloads).
Virtual machines are for testing and silly little one off instances of something that some department 'needed' on a 'server' that gets used by 3 people, twice a year, and thats only until next month when they've forgot about it completely.
VMs are for building an infrastructure that's easier to manage and scale than one built on physical hardware. If you're only using VMs the way you describe, you are Doing It Wrong.
In a lot of organizations it is cheaper to pay for a 24/7/365 MySQL expert from the local community. I know a guy who does just that, signs up businesses, gives them the cell phone number, and is prepared at any time to help them.
What does he do if two of them call at the same time ?
Circling above ? The chopper was a couple of *kilometres* away. They could probably barely even *hear* it.
The majority of wells show rising levels of pollutant, and the water tables in many places shows a consistent downward trend because we've been using it for irrigation and industrial processes -- and it's not an entirely closed system.
Where ?
But if we run out of drinking water because our wells and lakes are filled with poisons, all the oil in the world won't save our sorry asses. Every human being that ever existed needed water to survive -- daily. It's only the past few iterations of humanity that have needed it, and I'm certain we'll survive as a species without it, and without having to look at making large sections of land uninhabitable or reducing the population to do so. Water is serious business. Oil is a plaything for scientists and engineers to work out an alternative for. There is no alternative for water.
Nor is there any risk of running out of it. Ever. It covers 80% of the surface of the Earth. It falls out of the sky. You piss out 1-2L every day. It can be made potable by applying energy. We successfully build *cities* in the middle of the freakin' desert without even making _basic_ attempts at water conservation and reclamation.
We are not going to run out of water. Worst case is we need to build lots of pipelines, or relocate to places where precipitation is high.
The choice isn't that simple. LVM comes with its own complications, including a tendency to get volume offsets "wrong" so the file system data doesn't align nicely to RAID stripes. This is not good for performance.
This is only going to happen if your partition is misaligned, in which case whether or not you're using LVM is irrelevant.
The solution is to either a) use whole-disk PVs (preferable), or b) create properly aligned partitions.
Also, LVM has only recently acquired barrier support, and the combination of no barriers + write cache can be quite dangerous if power is lost. Even battery backed cache won't save you if you use a journalling file system (and everybody does these days) because request ordering isn't guaranteed.
Again, nothing to do with LVM, such a failure will affect any system.
And ... why would you do this?
To speed up and simplify restores.
The vast majority of "need a backup" situations are end users accidentally deleting things. This means a simple self-service restore capability that keeps, say, 1-2 weeks worth of hourly snapshots, will cover 80%-90% of restorations in a very short period of time with nearly no administrative overhead.
You can't do this with LVM (well, you can try, but performance is going to be dismal and it's going to be a nightmare to configure and manage).
Now, let's assume that we had to switch to desalination and purification of potable water in this country.
Why would you need to ? What is the basis of your assumption ?
Still think desalination is "easy" ?
It's a helluva lot easier than making more oil.
Oil. Even in the *worst* case water shortage is ultimately only a transportation issue. We really are going to run out of oil one day.
And The One would, of course, control everything, and thus be an archetypical instance of a totalitarian government. In fact wasn't that the goal of fascism?
Yes. Hence why removing laws and regulations to prevent that would be a bad thing.
And that's just fine, since any specialized version you might make is also in public domain.
Except for when it's tied down with hardware dongles, DRM, and phone-home schemes. Oh, and you can't get the source code any more, either.
GPL exists to hack around copyright law.
The GPL exists to ensure source code is available to anyone whom binaries are distributed to. It uses Copyright to achieve that goal.
No Copyright == no GPL == no source code available with distributed binaries. If the "source" part of "open source" is important to you, you need Copyright.
Upfront costs are one off costs, and if software was to start expensive and rapidly reduce to near zero cost as the initial costs were covered that would make a lot more sense. You may only have used 10 hours of support time, but you have support time available to you should you need it. Sure they make a profit, but they still need to have support staff on hand to answer your calls. Just because you aren't calling right now doesn't mean the call centre staff can pack up and go home.
No, the costs of initial and ongoing development are amortised across the expected lifetime of the product, and the "up front" pricing set accordingly.
You may only have used 10 hours of support time, but you have support time available to you should you need it. Sure they make a profit, but they still need to have support staff on hand to answer your calls.
Actually it's the updates I consider myself paying for - which is essentially the same situation as proprietary software, only a (higher) annual expense rather than a once-every-few-years cost.
Just because you aren't calling right now doesn't mean the call centre staff can pack up and go home.
Just because $SOFTWARE has been released doesn't mean all those developers can pack up and go home.
The issue is that proprietary software allows ridiculous profit margins (close to 100% since the software costs nothing to distribute and economies of scale are pretty much linear since the upfront costs remain the same regardless of volume)...
That the upfront costs remain the same does not mean they disappear. The proprietary software still needs to make up for its development cost before it's profitable.
You also assume said proprietary software is not maintained and updated after its initial sale (thus incurring more expense), which is patently false.
Open source vendors are unable to rip their customers off by selling zero cost goods at ridiculous markups because if they did someone else could come along and offer the same code for a cheaper price, instead they must make their money selling services... Services have a constant ongoing cost to actually provide the service, and these costs increase as you provide service to more customers.
We pay RH somewhere in the ballpark of fifty grand *per year* for our RHEL licenses. In the last 5 years, we've used maybe 10 hours of support time. I'm pretty sure they've made a lot of profit off us.
So can anyone name any large close source software companies that have started up rather recently that are billion dollar companies? I can't personally think of any. Can anyone else?
VMware.
I expect several of the computer game companies would also make the cut, though acquisitions and such might make that hard to figure out.
Anyway, if copyright laws didn't exist for software? Well, you'd see companies like Microsoft fall apart and companies like Red Hat thrive.
The GPL is utterly reliant on copyright to function as it does. Without it, the most useful GPLed code would be "proprietarised" in short order. Everything GPL proponents criticise the BSDL for would be applicable to GPLed software.
Because the business model would shift from protecting your source code through litigation to making it available for free since that would be the only way to effectively combat piracy.
No, the business model would shift to more effective piracy deterrents like hardware dongles and better DRM.
There is very little piracy of source code, most of it is binary distributions. Companies would probably lock their source code repos up a little tighter, and segregate them more so fewer (if any) employees had access to everything, but most big vendors will be doing this already. The removal of copyright would hurt open source (particularly GPLed) software far more - because while proprietary software vendors are used to dealing with piracy, OSS vendors always work under the assumption that no other OSS vendor will be able to gain a meaningful software-based competitive advantage.
Many businesses that reach billions of dollars in revenue often rely on government contracts and monopoly protection--patent law being the biggest of these. Without government interference in the economy businesses would probably be less likely to hit "billionaire" status. I don't doubt that there would still be some, just not as many.
More likely there'd be just one or two, at trillionaire status, and everything else would be subsidiaries of them.
In the open source world this is (to some extent) playing out.
Open Source software (at least the GPL or similar, which is typically what people mean) is utterly dependent on the "government monopoly protection" of Copyright. Without that, you don't have "Open Source", you have "Public Domain".
The fact is, the only way you would know that your business is serving people is if it makes a profit. Loss-making enterprises mean that there are better uses for your capital.
That's an awfully circular sounding definition.
You can't do that in Democrat-run states (like Maryland) either, so how are they any better?
I'm not making any commentary on whether the Democrats are better, I'm pointing out that the Republicans *clearly* are not interested in "maximum individual freedom" when gender discrimination is a fundamental and ingrained part of their platform.
You still haven't explained why people getting stuff for free because their friends gave it to them, and people getting stuff for free because their friends copied it for them, are different in terms of "ripping off the entertainment you want so you can have it with no money involved".
Republicans support a libertarian philosophy (albeit not as extreme as the actual LP) for maximum individual freedom.
Indeed. So much freedom you can't even get married to someone if they happen to be the same gender.