- Best way to use enum with classes - 25 Updates
| David Brown <david.brown@hesbynett.no>: Oct 27 08:41PM +0200 On 27/10/16 17:11, Jerry Stuckle wrote: >>> Give it up, David. The hole you're digging just keeps getting deeper. I will give up this "discussion" - it is just too tedious. I thought perhaps there might be a limit to how many things you could get wrong, but it seems not. (I knew there was no limit to how unpleasant and rude you would be.) It is clear to everyone that you know only little about RAID (or almost anything else you talk about), despite your claims of expertise. I say that you know a little - it would take very bad luck to guess incorrectly as consistently as you do, so I assume that you mix your bad guesses with intentional rubbish. Almost everything you say is in direct contradiction to widely available knowledge - anyone reading this thread who does not know you from before, and might be tempted to give you some credibility, will be able to check your "facts" with a few minutes web search. I am going to stop here, in the full knowledge that your other latest posts are also wrong, before I start using words that would make dear Wood Brian get his knickers in a twist. |
| Ian Collins <ian-news@hotmail.com>: Oct 28 07:53AM +1300 On 10/28/16 03:31 AM, Jerry Stuckle wrote: >>> RAID is the same as hardware RAID. >> Which is why David didn't. > You have already proven you can't read. Here's another instance. Time to end this. This snippet sums you in your own words. Too much of your twaddle is bad for the digestion. -- Ian |
| Melzzzzz <mel@zzzzz.com>: Oct 27 09:56PM +0200 On Thu, 27 Oct 2016 20:43:35 +0100 > > Time to end this. This snippet sums you in your own words. Too much > > of your twaddle is bad for the digestion. > Praise the f**king Lord, as Brian would very nearly almost say. There is no need to mask 'fucking' as there are no moderation bots on usenet... -- press any key to continue or any other to quit |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 26 11:22PM -0400 On 10/26/2016 2:27 PM, Chris Vine wrote: > I am glad that you appear to have understood it ... except that it also > appears below that you didn't. Stick to this one and you are doing > fine. Oh, I understand it perfectly. > with a given resistance, and (b) if there is no reactance in the circuit > (capacitive or inductive) - in other words, that the system is in a > steady state. That is correct. Ohms law is one of the basic laws of physics. Voltage is the difference in charges between two points. But if you have an infinite resistance, you can have no charge difference because there is no relationship between the points. > No. It is a source of EMF measured in volts (this time really by > definition). It *may* cause a current to flow in a circuit and so do > work, and will do so if a conducting medium is connected to it. Once again, false. A single electrode on a battery has no voltage. Only when you have a connection between the anode and the cathode can you have a difference in voltage. And that difference causes a current to flow. > The pressure in your water supply pipe does not go to zero when you > turn your tap off. The pressure is there all the time, courtesy of your > supply company. Only because on the other side of the tap you have a lower pressure. You have a connection between the two. But once again you can only measure that difference by measuring the difference between pressure in the water line and pressure external to the water line. >> connected. It is an extreme example to show the fallacy of your >> logic. > It shows no such thing. Then you can prove how connecting a voltmeter to just one electrode and placing the other electrode on the moon will show a voltage. It should be simple. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Christian Gollwitzer <auriocus@gmx.de>: Oct 27 08:55AM +0200 Am 27.10.16 um 05:13 schrieb Jerry Stuckle: > I have. The difference is I have used hardware RAID controllers - > unlike you. Of course, the other difference is I use my computers for > more than playing FreeCell. It seems you use it mainly to write excessive Usenet posts. Incidentally, the game you continue to refer to is from the early 90s. I think you should patent the time machine you are using. Christian |
| David Brown <david.brown@hesbynett.no>: Oct 27 01:42PM +0200 On 27/10/16 05:29, Jerry Stuckle wrote: >> drop-off of the electric field, and the potential is the integral of >> that (in which case I understand the 1/r factor)? > From a point source, it is 1/r². As Christian confirmed (since he knows the physics better than I), that is the drop-off for the electric field, not the potential. > From a surface, it is 1/r. But then > you should know that - after all, you're an expert in everything. I know that when you have an inverse square law (such as gravity, or in this case the electric field), then the drop-off from a point is 1/r². The drop-off from a /line/ (infinitely long wire) is 1/r. The drop-off from a surface - an infinitely extended plane, is 1. There is /no/ drop-off. (Of course in reality you don't have infinitely big lines or planes, and there is absorption in the media, etc.) >> I suppose the potential phi of an object is the energy it takes to move >> one Coulomb of charge from an infinite distance to the neutral object? > There is no such thing as a "neutral object". An electrically neutral object is one where the positive charges balance the negative charges. Most of the stuff you find around you is neutral (to a fair degree of accuracy, anyway). > And it would take an > infinite amount of energy to move anything, including a charge, an > infinite distance. You do realise that such descriptions are mathematical conveniences, not suggestions for practical physics experiments? To make them, you disregard loses, assume infinite time, an otherwise empty universe, etc. The energy needed in such circumstances is not infinite - it is defined, can be calculated, and provides a very useful number. It is like gravitational escape velocity. The escape velocity for the earth is about 25,000 mph at the surface. That means that if there were no other influences involved, if you launched from the surface at that speed then you would continue moving away without ever stopping and being pulled back to earth. This is referred to as being able to reach a point of infinity. |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 26 10:48PM -0400 On 10/26/2016 7:36 PM, David Brown wrote: > RAID is not a type of device or hardware, it is a way of organising > data. Whether you do it with dedicated hardware or pure software is > irrelevant. Only an incompetent idiot like you would even try to claim that software RAID is the same as hardware RAID. > raid, is its flexibility and its options. I didn't say you would be > likely to change these options in an existing system - it is possible, > and occasionally useful, but not common. And it's ability to corrupt the entire array due to a virus. But a COMPETENT person wouldn't even consider the need to change these options, because he or she would make the right decision in the first place. Just another straw man argument by an incompetent troll. > Read what I wrote - "I /hope/ you aren't trying to say...". Your > writing was unclear - and now you have confirmed what you were /not/ > saying. My writing was perfectly clear. I can't help it if your lack of understanding of computers and the English language doesn't allow you to understand it. But that's just like an incompetent troll. > Now, what /were/ you saying when you wrote "and small users sometimes > use Linux" ? Was it just another empty statement? Exactly what I said. Nothing more, nothing less. But I know you try to put words into my mouth. It doesn't work, troll. > functionality. But that is just like any other server - if you want the > server to be efficient and reliable, you use it for the functions and > programs that are needed for that server, and not for other purposes. First of all, it can be in ROM, also. There are still systems out there with their code in ROM - typically high security systems where the software *must* not be changed. But you've never worked on any critical systems - or even moderately secure systems, so you have no idea. And no, these servers are designed so that other software *cannot* run on them and potentially corrupt the disks. Not that you would understand the concept - it's way past your level of intelligence. > idea how it should work before starting the design? I bet even you have > a fair idea how a normal car engine works - but I also bet you have > never designed one. You can't even understand how something works, as you have repeatedly proven. And no, as any hardware designer will tell you - you can't really understand how something works until you have designed it. And only you would consider a car engine to be similar to a RAID device. Just more proof you don't understand RAID. > though I have little need for complicated RAID setups in practice. I > enjoy the mathematics involved in multi-parity raid. It is weird to do > maths for fun, I know, but there we have it. You can find the theory as interesting as you like. But you don't understand even the basics of a RAID system - only how to use one. And even that understanding is demonstratably limited. > No, if your raid controller card died you would need to get a > replacement raid controller card of the same (or very similar) type. > That is not the same thing as getting another computer. Or you just get another RAID device and move your disks to that device. No problem. > BSD, or FreeBSD, or whatever) that you had on the old system, or you > install a newer version if you want. It will still work fine with the > disks. For the same reason you are using an old or unusual RAID device. I'm comparing apples to apples. But I know you can't do that, because it destroys your arguments. And the old version of Linux may no longer be available, and new versions may not run on the old device. Just like an old controller card may not be available and a new controller card may not work in the RAID device. > outside forces, then you will not expect to have any serious risk of > malware infections within the working lifetime of the server. That does > not mean that it cannot happen - it means it is highly unlikely. Not in any of your systems, it isn't. But a properly designed RAID device is immune to such problems because other code *never* runs on the RAID device. >> permissions. > Have you ever actually /used/ any sort of raid system? I mean, have you > ever configured a raid system and installed an OS on a PC that uses raid? I've been using RAID since the 80's. But I've never installed an OS on a PC that uses RAID. They are too insecure. > configuring the hardware raid sets, it could even screw with them (it is > highly unlikely that it would bother, of course - it's easy enough to > destroy the data with a total disregard for any underlying raid). I know how it appears. And I know that some file systems have RAID emulation built into them. But you once again show your lack of knowledge when you claim that a hardware RAID will execute a virus. It can't, because there is no way for any external code to be executed on the RAID device. It can't run partd. or anything else. A virus on a user computer can corrupt individual files the user has permission to write. But there is no way it can corrupt other files or the disks, as can happen with software emulation. And you claim that RAID emulation can be done on multiple virtual partitions on a disk. But that violates your earlier statement that one of the advantages of RAID is to be speed up access by spreading the data over multiple physical disks. Which is it? Oh, I forgot. It's whatever argument is convenient for you at the time. You change your tune more often than some people change their shorts. But that's what a troll does. >> You can do the same on RAID, if you wish. > You can do it if you have a filesystem or volume manager that supports > fast snapshots, or do it slowly without such support. Sure. And you can fly to the moon if you have a big enough rocket. But RAID can do it automatically, with no speed reduction. > might have. > But you only have backups if you realise that your hardware raid card is > not a backup solution as you have repeatedly claimed. Nope, if you have RAID emulation a virus can corrupt the entire disk. But the hardware RAID card can easily be a backup solution. The most obvious way is by configuring it for RAID 1. But once again you show your ignorance by your statements. Give it up, David. The hole you're digging just keeps getting deeper. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 27 10:28AM -0400 On 10/26/2016 10:43 PM, Ian Collins wrote: >>> http://wiki.qemu.org/Main_Page >> So is RAID emulation. No difference. > I guess the real world and Jerry world will have to disagree on that one. The real world and my world agree on this. But the "expert" who a few days ago admitted he knew little about RAID disagrees. ROFLMAO! > To confuse you further: ZFS can emulate Redundant Array of Independent > Disks by using disk files as emulated disks. So what? Can't you even stay on topic? No - when you can't counter an argument, you change the subject. Typical of a troll. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 27 11:19AM -0400 On 10/27/2016 2:55 AM, Christian Gollwitzer wrote: > Incidentally, the game you continue to refer to is from the early 90s. I > think you should patent the time machine you are using. > Christian Actually, it's from much earlier than that. But it's still above David's level of competence. And I write the posts in an effort to correct many of the misunderstanding some people have about topics they obviously have no real idea on, except what they might have read in Wikipedia. Unfortunately, it's like trying to teach a pig to sing. But maybe some others will learn. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Ian Collins <ian-news@hotmail.com>: Oct 27 08:04AM +1300 On 10/27/16 12:47 AM, Jerry Stuckle wrote: > No, people who need high reliability backup use hardware RAID. RAID != backup. > includes U.S. government and military, major corporations and others. > And yes, they accept whatever IBM or Dell tell them to because those > people know a lot more about it than you do. So Oracle and its customers (who include the U.S. government) aren't "people who need high reliability" or big companies? > Or are you claiming you know more about it than the big companies with > thousands of experts do? The big companies who use Oracle? -- Ian |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 28 05:27PM -0400 On 10/27/2016 7:32 PM, Chris M. Thomasson wrote: > http://webpages.charter.net/appcore/fractal/field > You can make dipoles, field lines, equipotential lines... > ;^) Which has absolutely nothing to do with electrical charges. But once again you try to change the subject. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 27 11:21AM -0400 On 10/26/2016 11:25 PM, Ian Collins wrote: > Where did I say I know very little about RAID? Come on, show the quote > and prove me wrong. > I've been in the storage business for a very long time. Early in this thread you admitted you didn't know much about RAID configurations. Or was that another Ian Collins? And according to your claims, you've been in EVERY business for a very long time. Your lack of knowledge belies that. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Christian Gollwitzer <auriocus@gmx.de>: Oct 27 09:44AM +0200 Am 26.10.16 um 23:57 schrieb David Brown: > Is it not 1/r² drop-off from a point charge? Or would that be the > drop-off of the electric field, and the potential is the integral of > that (in which case I understand the 1/r factor)? Yes, exactly. For a single isolated static point charge at the origin, the potential is phi(r) = q/(4*pi*epsilon0) *1/r. The electric field is the gradient of this potential E=-grad phi = q/(4pi*epislon0) * 1/r^2 *r^, where r^ is a unit vector pointing from the origin. A battery can be considered an electric dipole, for which indeed the potential drops off at a higher rate, phi~1/r^2 and E~1/r^3. See https://en.wikipedia.org/wiki/Electric_dipole_moment for plots and formulae. > I suppose the potential phi of an object is the energy it takes to move > one Coulomb of charge from an infinite distance to the neutral object? Yes, the potential is usually defined in this way that at infinity it is zero (because for a finite amount of charges, it always drops off), but in general it is defined only up to constant. The energy difference in moving a test charge q from x1 to x2 is q*(phi(x2)-phi(x1)). The endpoint can be anywhere in space, not necessarily on the surface of the object. Christian |
| David Brown <david.brown@hesbynett.no>: Oct 27 01:58AM +0200 On 27/10/16 00:08, Jerry Stuckle wrote: >> more slowly) even if you have a disk failure. That is what the R for >> /redundancy/ means. > Yes, which is a form of backup. OK, so you /are/ using the wrong words. You are making up your own terms, just like your "true raid" and "emulated raid" terms. For future reference, if you want to communicate with people about technical issues, it makes sense to use the standard terms. You might like to start by googling for "backup" and "raid" - or perhaps go straight to Wikipedia. >> if you simply need to refer to older versions for some reason. > Two copies of a file IS backup. Even copying a file to a different > directory is a well-known method of backing up the file. To be a backup, you need two /logical/ copies of a file. Two physical copies of the data block is redundancy, not a backup. A COW logical copy (such as you get with LVM, BTRFS or ZFS snapshots) is a backup (without redundancy) even though there is only one physical copy of the data - you can view them as independent files, change one while having the old backup copy remain safe, etc. Copying a file to a different directory is a simple backup, yes. >> replication of data. > Yes, and what do you think "continuous off-site replication of data" is, > if it's not off-site backup. It is not a backup, because it is not logically independent. Any changes you make to the on-site copy are quickly (depending on network speed) copied to the off-site copy. If normal usage changes your copies, they are not backups! Also note that this sort of continuous off-site replication is not a common usage, and it is certainly not supported by any hardware raid system. >> data is needed in an application - the data has to pass through the cpu >> anyway. > Your proof? You don't have it, because once again you are full of crap. Do your own research. It should not take long to google. > Even with two disks you are halving the amount of data going through the > CPU. And it is much more than just the data. It includes commands to > two (or more) drives, additional wait times for seeks, etc. So what? Double a small amount is still a small amount. On my servers I see a few percent cpu time used for the raid1 sets. Even with raid6, the cpu time is generally irrelevant. With a half-decent cpu, it is faster to use transparent compression on your filesystems - even with the cpu time needed for compression and decompression, the throughput is faster than reading straight from a hard disk. And compression and decompression is a good deal more cpu intensive than even raid6. > Current controllers use processors just as fast as modern computers. In > fact, some are faster because they can use RISC processors and code > optimized to those processors. No, they are not faster. And they have significant latencies compared to direct host access of SSD's. >> The processor has more than enough cycles to spare. And if it doesn't, >> it is more cost-effective to buy more or faster cpus. > Of course it does, when all you use it for is to play FreeCell. Maybe you play Freecell on your servers, but most of us do not. >> reference. > And "may have meant"? Based on what facts? At least you admitted you > don't have any facts - just an opinion. I never claimed to have facts here - just my interpretation. And I am sure Ian will let me know if I guessed incorrectly. > Sure, they use RAID. And as I said earlier, RAID uses software. But > these devices are dedicated to backups - no user code is running on > them. A huge difference. In what possible way is it different? You run appropriate server systems on your servers - you don't let any user log in and run whatever "user code" they fancy. If your machine is primarily a file server, then you run file serving services on it. If it is some sort of application server, then you run those services on it. And if you really have services that demand maximal cpu time, then you are likely to want the fastest raid types - raid1 and raid10, which have completely negligible cpu costs. |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 26 10:55PM -0400 On 10/26/2016 7:48 PM, Ian Collins wrote: >> I'm claiming that the major use of RAID around the world is for reliable >> data backup. > The major use of RAID around the world is for reliable data storage. So that is different from your previous claim that the major use is for fast access? Which is it? And reliable data storage requires data backup. >> Yes, which is a form of backup. > Nope. That's a common mistake naive admins make until their machine > goes pop or their RAID card shits its self. No, it is a common mistake people like YOU make that a backup is only manual, only happens on occasion, and can be out of date. High reliability systems run constant backups using RAID and can switch to the backup at any time with no loss of data. But you wouldn't know what a low reliability system is - much less high reliability, as you have repeatedly demonstrated. >> Two copies of a file IS backup. Even copying a file to a different >> directory is a well-known method of backing up the file. > Nope, its redundancy. It is also backup. See above. >> Yes, and what do you think "continuous off-site replication of data" is, >> if it's not off-site backup. > Redundancy. See above. >> them. A huge difference. > Indeed, they are not dedicated to backups, they are dedicated to > redundancy. What you call redundancy is just another form of backup. Not that you would understand the concept, because you have never worked on a system requiring high reliability. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 26 10:50PM -0400 On 10/26/2016 6:34 PM, Ian Collins wrote: >> No one ever said it was. However, you have to try to put words into >> other peoples' mouths to try to make a point. > You keep conflating the terms. Nope. You just keep trying to put words in other peoples' mouths to try to support your idiotic claims. >>> "people who need high reliability" or big companies? >> You don't know the difference there, either. > Difference between what? I rest my case. >> Yes, and the big companies who use real RAID devices for their critical >> reliability systems. > Oracle engineered systems all use ZFS. Oh yes? You know every Oracle system? Quite a grand statement from someone who just a couple of days ago admitted he knew very little about RAID - but now you're an expert in everything RAID - including every product Oracle has. Want to try again, troll? -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 26 11:27PM -0400 On 10/26/2016 12:45 PM, Christian Gollwitzer wrote: > between one of it's connectors and an infinitely distant point? The > answer is U/2, half of the battery voltage. > Christian Christian, Yes and now. Phi is a theoretical voltage, quite useful for modeling ideal systems. But in real life, systems are never ideal, and to measure that potential difference upsets the theoretical model. Just like modeling a battery in electrodynamics assumes an internal resistance of 0 ohms, which is never the case in real life. But it simplifies the modeling, and is often small enough that it can be ignored in most situations. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Ian Collins <ian-news@hotmail.com>: Oct 27 03:43PM +1300 On 10/27/16 03:27 PM, Jerry Stuckle wrote: >> "QEMU is a generic and open source machine emulator and virtualizer." >> http://wiki.qemu.org/Main_Page > So is RAID emulation. No difference. I guess the real world and Jerry world will have to disagree on that one. To confuse you further: ZFS can emulate Redundant Array of Independent Disks by using disk files as emulated disks. -- Ian |
| Gareth Owen <gwowen@gmail.com>: Oct 27 08:43PM +0100 > Time to end this. This snippet sums you in your own words. Too much > of your twaddle is bad for the digestion. Praise the f**king Lord, as Brian would very nearly almost say. |
| "Chris M. Thomasson" <invalid@invalid.invalid>: Oct 29 03:15PM -0700 On 10/28/2016 2:28 PM, Jerry Stuckle wrote: >> :^) > Again having absolutely nothing to do with electrical charges. But it > seems trying to change the topic is a favorite pastime of yours. Well, you can create an image resembling a dipole... But, so what, wrt changing topic, yes, I do suffer from that problem! Sorry. |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 27 11:30AM -0400 On 10/27/2016 5:14 AM, Chris Vine wrote: >> have an infinite resistance, you can have no charge difference >> because there is no relationship between the points. > You are clueless. Not at all. Very basic - something I learned back in 5th grade science class. See my previous post for an example. You obviously have no idea. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| David Brown <david.brown@hesbynett.no>: Oct 27 01:12PM +0200 On 27/10/16 04:48, Jerry Stuckle wrote: >> irrelevant. > Only an incompetent idiot like you would even try to claim that software > RAID is the same as hardware RAID. If "hardware RAID" and "software RAID" were the same thing, there would be no need for two terms, would there? But both are RAID - in the same way that a "blue car" is not the same thing as a "red car", but they are both cars. And for a good many uses, it does not matter much which type of RAID you use. To the filesystem, it is usually all the same - it's just a big block device. (Some filesystems can use knowledge of the underlying geometry of the block device to improve performance, but that's a bit too technical for you.) >> likely to change these options in an existing system - it is possible, >> and occasionally useful, but not common. > And it's ability to corrupt the entire array due to a virus. When you invent an irrelevant and incorrect straw man in one post, you don't need to keep it up in other posts. > But a > COMPETENT person wouldn't even consider the need to change these > options, because he or she would make the right decision in the first place. Like any competent person, I would try to make the right (or at least, a "good enough") decision in the first place. But like any competent person, I know that circumstances may change in the future - and that may mean it is worth changing the RAID layout of your disks. > with their code in ROM - typically high security systems where the > software *must* not be changed. But you've never worked on any critical > systems - or even moderately secure systems, so you have no idea. I know where and why ROMs are used. And they are not used to store the software in a NAS. > And no, these servers are designed so that other software *cannot* run > on them and potentially corrupt the disks. Not that you would > understand the concept - it's way past your level of intelligence. Some let you run other software, some do not. You have this weird idea about "user code" and "viruses" destroying your disks, partitions or filesystems and that somehow hardware raid is immune to it but it is a common problem for software raid. It is simply utter nonsense, with not the slightest basis in reality. Your server OS sees hardware raid volumes and software raid volumes at the same level - there is a big logical volume which can be partitioned, and filesystems can be created on those partitions. Any "user code" or "virus" which can destroy the partitions on a software raid can do exactly the same thing on hardware raid. > You can't even understand how something works, as you have repeatedly > proven. And no, as any hardware designer will tell you - you can't > really understand how something works until you have designed it. Remind me never to buy anything /you/ have designed. > And > only you would consider a car engine to be similar to a RAID device. > Just more proof you don't understand RAID. Does the word "analogy" mean anything to you? >> maths for fun, I know, but there we have it. > You can find the theory as interesting as you like. But you don't > understand even the basics of a RAID system - only how to use one. So now you are ready to accept that I can /use/ a RAID system? I guess that is a big step forward. And you are happy to accept that I understand the mathematics behind RAID? But somehow you also think I can't understand even the basics of a RAID system? > For the same reason you are using an old or unusual RAID device. I'm > comparing apples to apples. But I know you can't do that, because it > destroys your arguments. /All/ raid cards as unusual. There is no real standard (some cards support DDF standards, to some extent). There are a few big players who make the controller chips and the basic card design, and there are a dozen companies who re-badge them. There may be some interoperability between them, but there may not - even if the chips are the same. You might have an IBM badged RAID card that will only talk to IBM badged disks, a Dell RAID card that will only talk to Dell disks, and an HP RAID card that will only talk to HP disks. It could be that all the disks are Toshiba inside, and all the cards have the same LSI controller. But "to make sure everything works together" or "by the terms of the service contract" or "for a consistent user experience", everything is vendor-locked. > And the old version of Linux may no longer be available, That would only be if you don't choose your Linux distribution sensibly, /and/ you don't keep track of your installation media. For most distros, old versions are easily available and you can install as much as you want (commercial service contracts are, of course, time limited). > and new > versions may not run on the old device. What "old device"? You are putting the old disk into a new machine, you are not using old devices. >> ever configured a raid system and installed an OS on a PC that uses raid? > I've been using RAID since the 80's. But I've never installed an OS on > a PC that uses RAID. They are too insecure. Which means that your actual real /usage/ of RAID is limited to a few basic hardware RAID systems, and your experience with installation, configuration and setup is zero, as is your experience with partitioning, filesystem creation, OS installation, etc. That explains some of your outdated ideas, and the gaping holes in your knowledge. You are a fine example of how a little knowledge can be a dangerous thing - you know a little about one corner of a subject, and think you are an expert on it all. > user computer can corrupt individual files the user has permission to > write. But there is no way it can corrupt other files or the disks, as > can happen with software emulation. A hardware RAID card (we are talking about hardware RAID cards here, not NAS systems) presents its raid sets as virtual disks to the OS running on the server itself. The /server/ runs whatever code you want - and if you have a security breach, it may also run code that you /don't/ want. No one suggests that the virus is running on the RAID card, any more than they suggest that the virus would run on the disk controller processors on the hard disks. > partitions on a disk. But that violates your earlier statement that one > of the advantages of RAID is to be speed up access by spreading the data > over multiple physical disks. Which is it? I said that the RAID layer - whether hardware or software - presents the RAID sets as virtual disks to the OS. If you have four 1TB harddisks set up as RAID-5, the OS sees a single 3 TB virtual disk for partitioning and filesystems. It is also correct that software RAID can use multiple partitions on a disk independently as block devices for the RAID set. It can also use other RAID sets as block devices, for layered RAID setups. Different partitions can be attached in RAID setups in different ways to provide different balances between speed, redundancy, and space efficiency. A key advantage of software raid is you have the flexibility to choose what makes sense to /you/. As a simple example, you might have four 1 TB disks, sda through sdd. You could partition them as: 1: 512 MB 2: 4 GB 3: 4 GB 4: about 1 TB (the rest of the disk) The four sd?1 partitions you set as a four-way RAID-1 mirror, md0 at 512 MB. You use this for /boot and the bootloader. Then the system will always be able to boot from any disk, and the bootloader does not have to understand RAID formats because the layout is identical to a non-RAID layout. The four sd?2 partitions you set as swap. There is no redundancy (usually you don't need that for swap, but it's possible if you want it). The OS will automatically handle them as though it were a 16 GB RAID-0 array - there is no need for an explicit md raid device. The four sd?3 partitions you set as a RAID-10 array md1, and use the 8 GB for root for the OS. This gives you fast, low-latency access to all the small files of the OS, with redundancy. The four sd?4 partitions you set as a RAID-5 array md2 for greater space efficiency (but less speed), with 3 TB available for /home for user files. /Flexibility/ is the key property of software raid. With the same RAID setup, it will usually be faster than hardware raid - but that is only one if its benefits. >> fast snapshots, or do it slowly without such support. > Sure. And you can fly to the moon if you have a big enough rocket. But > RAID can do it automatically, with no speed reduction. No, RAID cannot do snapshots. Not hardware RAID, not software RAID - it is not part of what RAID is or does. > But the hardware RAID card can easily be a backup solution. The most > obvious way is by configuring it for RAID 1. > But once again you show your ignorance by your statements. Attempting proof by repeated assertion, yet again? |
| "Chris M. Thomasson" <invalid@invalid.invalid>: Oct 29 03:15PM -0700 On 10/28/2016 2:27 PM, Jerry Stuckle wrote: >> ;^) > Which has absolutely nothing to do with electrical charges. But once > again you try to change the subject. Sorry. |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 27 11:17AM -0400 On 10/26/2016 11:07 PM, Ian Collins wrote: >> So that is different from your previous claim that the major use is for >> fast access? > I made no such claim. You didn't? I guess there's another Ian Collins here, then. >> And reliable data storage requires data backup. > Indeed. The two are not the same. Not according to you. Data backup means a second (or third or whatever number) of copies of the data. This can be static backups - what you might do once a year or so, if forced by your manager to do so. But you can also have copies updated when the original file is updated - as in a RAID 1 setup. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| Jerry Stuckle <jstucklex@attglobal.net>: Oct 27 11:37AM -0400 On 10/27/2016 7:42 AM, David Brown wrote: >> From a point source, it is 1/r². > As Christian confirmed (since he knows the physics better than I), that > is the drop-off for the electric field, not the potential. That is correct. And there is a direct relationship between the potential and the electric field. > drop-off. > (Of course in reality you don't have infinitely big lines or planes, and > there is absorption in the media, etc.) Incorrect again. The drop off from a wire is 1/r² perpendicular to the wire, and no drop off parallel to the wire. The drop off perpendicular to a surface is 1/r. > An electrically neutral object is one where the positive charges balance > the negative charges. Most of the stuff you find around you is neutral > (to a fair degree of accuracy, anyway). There is no such thing as a "neutral object". It is all relative. When you have a difference of 100V between two objects, you can claim one has a +100V charge or the other has a -100V charge. Or you could claim +40 and -60. Or anything else in between. It's all relative. > disregard loses, assume infinite time, an otherwise empty universe, etc. > The energy needed in such circumstances is not infinite - it is > defined, can be calculated, and provides a very useful number. Even as a mathematical convenience it would take an infinite amount of energy to move anything, including a charge, an infinite distance. > speed then you would continue moving away without ever stopping and > being pulled back to earth. This is referred to as being able to reach > a point of infinity. That is not the "point of infinity". That is simply Earth's escape velocity. And BTW - all it would do is escape Earth's gravitational field. It would still be orbiting the sun. -- ================== Remove the "x" from my email address Jerry Stuckle jstucklex@attglobal.net ================== |
| You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to comp.lang.c+++unsubscribe@googlegroups.com. |
No comments:
Post a Comment