How on Earth is RAID 5 "less real" than RAID 1 or 10 ?
I mean only that using RAID 5 your write performance always suffers due to parity calculations that have to be done on the data. What suffers even worse is read-modify-write, whenever existing data needs to be modified. Your RAID controller or kernel has to read the data into memory, modify it, then calculate parity on the entire modified block, then write back out to disk. Even if you only make a 1 byte modification to a file, the read-write-modify cycle takes a huge performance hit and parity must be calculated for the entire file.
For this reason, most DBAs don't consider RAID 5 useful for serious OLTP type databases.
Yeah--oddly though I'm getting better Bonnie results by using Linux RAID 5 than their hardware RAID 5. But it is possible (just) to stream full-size, full-framerate PAL video to 'em over NFS! (sustained 40MB/s). Anyway in software you can now do RAID 6:)
The reason why you're getting better RAID 5 results from software RAID vs. hardware RAID is because of the parity calculations involved with writing to a RAID 5 volume. On a hardware RAID setup, these are calculated on the RAID card itself, which probably has a 200 or 400 mhz. chip that does these calculations. Back when CPUs were only 400 mhz, this was great, because there was no load put on the CPU, and the RAID controller worked just as fast or faster than a software RAID setup. Now that CPUs are 3 ghz. +, there's no way a dedicated hardware RAID card can keep up, and unless you're running a huge load on the server, youv'e probably got 1 ghz. or so of free CPU bandwidth to burn for software RAID...
Want to see the performance really increase? Give up RAID 5 and go with a real RAID solution like RAID 1 or RAID 1+0.
Did companies using PowerPC based IBM servers manage to stay in business during the dot.bomb? Maybe the saying "nobody ever got fired for buying IBM" has some solid reasoning behind it.
I don't think your choice of Unix vendor can save your company if you have a failing business model. The companies that made it, like eBay, have a pretty healthy mix of Sun, HP, and probably Linux as well.
I'm curious, why do linux distributions typically have PowerPC releases always up to a generation behind when it would seem that HPPA and SPARC are likely a smaller installed base?
Is it that there are relatively few of the PPC base demand a Linux distribution when compared to what are mostly server-class CPU's and more likely to be running a Linux distro?
Well, I believe this is due to two things. First, you have the eBay phenomenon. Tons of Sun and HP hardware available for dirt cheap on eBay, and we're talking server class machines for a fraction of their dot.bomb retail prices. That, plus a lot of techs got "free" boxes when their dot.bombs went under and they just sort of "acquired" boxes that would otherwise be repo'd by the creditors.
I'm curious to know if this Sparc release of Gentoo actually frickin' works or not! I tried to install Gentoo 1.4 on a Sparc and it was a frickin nightmare. I did finally get it working (somewhat), but X barely worked, and most packages wouldn't emerge properly. I couldn't even think of getting KDE working on the thing as the emerge would just crap out on me. It seems that the Sparc maintainers don't really support or even maintain the product.
It's also not a very good sign when you go to the release website and there are no release notes available for Sparc or HPPA platforms... Gee, if they're missing the release notes, I wonder if they ever even released the software.
As usual, I'm betting this is just a "vapor" release on all platforms except for X86 and AMD64. Why does such a cool distro have to torment me so? If you can't release a decent Sparc distro, don't even waste my time announcing it. I'll just stick with Debian, which as far as I know is the only current distro that actually works on Sparc.
Let's see, what would I do with all that bandwidth:
Try to find a host OS with a TCP/IP stack that can properly utilize 1 gigabit ethernet, let alone 10 gigabits. Hint: It ain't Linux...
Try to find a storage solution that can read or write that fast. I'm thinking something like EMC with about 6-8 2 gigabit HBAs using Veritas DMP (dynamic multi-pathing).
Try to get all of the above, along with a 133 mhz. 64-bit PCI-X bus that still can't actually keep up with 10 gigabits of data. (133 mhz. 64-bit PCI-X is only about 1024 megabytes per second, not counting overhead).
The problem is, right now, the rest of the parts of a system just can't keep up with 10 gigabit ethernet. The only box that I would use that can handle that many I/O paths to storage (we're talking six to eight 64-bit 66 mhz. 2 gigabit FC host adapters) is a Sun Fire 6800 or something larger. The problem is, Sun doesn't yet support PCI-X, so now your 10 gig ethernet card is going to be limited to a 66 mhz. 64-bit PCI version, which will only transfer a maximum of 512 MB per second, not counting overhead. That is less than half of the available bandwidth of 10 Gig Ethernet.
You can forget about putting it in any Intel based system. There are not enough I/O busses and I/O controllers in even the beefiest Xeons or Opterons that can handle this much bandwidth (to disk).
Also, if your application doesn't need to write all of that data to disk, then how large is this dataset in memory that needs to be transferred at 10 gigabit speeds? If you had a server with 64 GB of memory, it could transfer it's entire memory set over 10 gigabit ethernet in less than 60 seconds.
A far better, and more economical solution, if you really need 10 gigabits of data throughput to the network, would be to use the same Sun server, and a product called Sun Trunking, which allows you to bond multiple gigabit ethernet interfaces together. You get all of the throughput you want, plus more fault tolerance. I've set it up before, and you can have a continuous ping going, across 4 connections, and pull 3 of those 4 connections and the ping keeps going, without even a dropped packet. It's really fault tolerant, and uses your existing switches, NICs, and hardware, without forcing you to upgrade your entire core switch architecture.
Doesn't anyone else besides me see a conspiracy theory in the making?
You're not alone at all. I got modded as flamebait for stating this in someone's journal. The DHS declaration last week that "polling places will be likely targets of terrorist attack" seems designed to instill fear in Joe Sixpack and keep him from voting our incumbent Republican government out of office.
The simple truth of the matter is that polling places are usually churches, schools, people's garages and homes. To think that somehow Al Qaeda is going to target the tens of thousands of small neighborhood polling places across the country is ludicrous.
The government that is in power will try whatever dirty tricks are necessary to win. They're not above it.
You know what is really amazing... The mayans had a very sophisticated calendar which was entirely based around the cycles of the Sun's magnetic fields. I heard an interesting speech by Maurice Cotterell, an engineer who heavily researched the Mayan calendar, and he predicts that in the year 2012, there will be some type of major magnetic field shift in the magnetic fields of the earth at this time. He doesn't believe the world will end then, but he believes that there could be major catastrophes because of this. If you get a chance, I recommend reading one of his books and deciding for yourself whether what he says is believable.
Maybe I can use this to track down the scumbags who send junk faxes to me at all hours of the night and morning, but whose numbers are listed only as "Out of Area". In fact, I bet this would be a handy tool for those who are trying to stop these asshats.
It won't be too long before these asshats figure out how to spoof on their own, then it will be just as bad as email spam where the headers give you no clue as to the actual source of the email (other than the sending MTA's IP address).
Kinda hard to surf the net at the speeds mentioned in the article when carriers like Sprint haven't made 3G fully functional. I am still only able browse the web at a mere 5k or so, 10k bursts if I'm lucky.
Really? I consistently get 160kbps with my Sprint Treo 300 connected through a custom PDANet driver to my laptop (Windows only, sadly). There is also a program called WirelessModem for the Mac, but it's so buggy it caused kernel panics every few hours or so (grey screen of death) on my PowerBook.
Sprint uses 1xRTT for data, who's theoretical maximum is (I believe) 224kbps. Of course, as with any wireless protocol, there is overhead. The reason you're seeing slower speeds is (I'm just guessing here) probably because you're connecting over Bluetooth. Although the Bluetooth spec supports speeds up to 768kbps, most BT phones serial port profile only supports 115200 bps (just like a regular serial port), so you're limited by the bandwidth of that connection. With a Sony Ericsson T610 I can only get about 5KB (kilobytes) per second, even when doing a Bluetooth file transfer, so it's pretty unlikely that I'd get any better than that using the Internet.
So, in conclusion, if you want 160kbps through your 1xRTT Sprint PCS connection, get a phone that supports a USB connection instead of Bluetooth. The difference is night and day.
You're a wonderful example of the ignorance which is so pervasive through this industry.
Nice... I just checked out your personal website, and I see you've bought the MS party line hook and sinker. Let's see, you're using a notoriously riddled with security holes CMS (Nuke), running on top of.NET (which isn't that bad in itself, except for the fact that MS seems to change dev tools every month now so if you want to keep porting your code, go right ahead), using a webserver (IIS) that's riddled with holes.
You still haven't rebutted my initial argument that user-level webservers will always be more secure than kernel-level webservers. Care to try?
Microsoft provided a patch for this IIS exploit back in April.
I should have clarified that statement. There's currently no client patch available for the millions of IE users that are most at risk of having keystroke loggers installed.
You're a wonderful example of the ignorance which is so pervasive through this industry.
And you sir, are a Microsoft shill. No doubt the majority of your income comes from foolish customers that are simply too ignorant to know that there are superior free solutions to the expensive MS solutions you recommend on a daily basis. Besides, all the support calls sure boost your billable hours. If I were in need of a corporate website, your company would be the last one I'd contact, for that very reason.
Cisco Press books are, without a duobt, the best technical manuals (from a manufacturer) that I have yet read.
I agree wholeheartedly. Especially Basam Hallabi's Internet Routing Architectures. (No affiliate link) This book taught me how to establish BGP routing policies, and is considered fundamental reading by almost anyone on NANOG.
Can tend to throw the beginning student off, especially considering the OSI model is not much more than an academic tool anyway, TCP/IP is were its at in the 'real world'.
It's good to know how the OSI model works, especially when troubleshooting tricky layer 2, 3, and 4 network issues. That being said, when I took the CCNA test they only had about 2 questions on it. They were something like "Which of the following are layers in the OSI model" or something really basic like that. Anyway, my instructor had a really good way to remember the OSI model that I still remember to this day (took the class in 1997):
All People Seem To Need Data Processing.
The letters in this saying correspond very nicely with the layers in the OSI model:
Application
Presentation
Session
Transport
Network
Data Link
Physical
Hope this helps those of you studying for your CCNAs (judging by enrollment numbers in these college courses, there's a lot of you.)
If you register at Lulu, the free download is a PDF.
Wow, Lulu.com is Slashdotted already... There must be a pretty huge demand for this book. I know I haven't been CCNA certified since version 1 of the test (they are up to 3 now), but I'm downloading the Word version because someday I'd like to update that certification.
... It's things like this that make me want to go back to the store. Of course when you talk rebates though, CompUSA is definitely the KING of rebates.
Are you and I thinking of the same store? CompUSA is the worst at returns. They have a 15% restocking fee on everything, even if it's in brand new condition in the original packaging. Let me tell you my CompUSA story:
A couple years ago, I was in the market for a PDA. I'd heard good things about the Handspring models, so I went in and purchased one (can't remember which one) for $300. It was pretty nice, but the next day I was reading Slashdot and saw an article on the Sharp Zaurus SL-5500. It ran Linux, had a ton of open source software available for it, and was only $100 more expensive than the Handspring. I had to have one. So, I immediately wiped the memory on my Handspring (put it back to factory defaults), boxed it back up in the original packaging, and took it to CompUSA to exchange it for a Zaurus. It was in brand new condition, barely used, no scratches, with the original receipt. I wanted to exchange it for the Zaurus, and I was willing to pay the $100 difference between the two models.
The manager tells me that he is going to charge me a 15% restocking fee on the Handspring to return it, even though I was buying a more expensive model at the same time! Fucking ridiculous. I sat there and argued with him for about 30 minutes. I showed him receipts proving that I had spent $2000 there in the last month, on everything from printers to computer accessories to software, and none of it returned. Not only that, I was paying full retail price for the Zaurus just because I wanted it right now. He wouldn't budge. I told him flat out that I would never shop at CompUSA again, and that he had lost a valuable customer. I returned the damn Handspring, -15% of course, and I've never shopped there since.
CompUSA doesn't deserve any of my business. They also have some of the worst prices in the industry. You can occasionally get a good deal there if you look for the rebates or loss leader items, but it's not worth the 15% restocking fee if you have any problems with it. Now I buy all of my computer hardware from NewEgg, and I've never looked back. Newegg is mail order, but they've got some of the best customer service I've ever seen. I had a memory DIMM fail once after 8 months of use, and all I had to do was request an RMA from their website, ship it to them, and the very next day they sent me a brand new replacement, which arrived in 2 days from FedEx. All in all I was only without the DIMM for a week. Same thing happened with a video card that stopped working after a few months. I got a brand new one which happened to be better than the one I returned, within a week. I guarantee you that even with the PSP or whatever purchase protection plan they have, you're going to wait a couple of weeks minimum to get your repaired merchandise back from CompUSA.
Never again will I shop at CompUSA. Best Buy is just as bad. If I had to pick one retailer to get "instant gratification" shopping, where you want the item right that second, it would be Circuit City, just because they aren't quite so draconian about returns and exchanges. They even have that 110% pricematch thing, so if you see it for cheaper next week in a Best Buy ad you end up getting a sweet deal.
Under this plan, I tried taking the stereo back because it wasn't functioning properly.
The moral of this story: Always purchase electronics with a credit card. If the merchant won't take it back and immediately exchange it for a working version of the same product (no 2 weeks waiting for repair, I'm talking about immediate exchange right there in the store), call your credit card company and dispute the charge. Works every time. The fact of the matter is that the credit laws protect you because you do not have to pay for goods or services that you never received, or were defective. The credit card company can't bill you for them and will issue a chargeback which means the merchant won't get their money either.
Also, if a merchant gets too many chargebacks in a certain period of time, the credit card company will revoke their merchant account, leaving them unable to accept credit cards period.
This is also a highly effective bargaining tool if the customer service rep is giving you a hassle about returning the item. Just tell them you will dispute the charge with your CC company if they don't accept the return right this second. They will immediately cave, because they know they don't have a chance of winning that battle.
A business is under no obligation legally to take back a returned item once sold, as long as it is not defective and was not sold under false pretenses. Businesses like BestBuy take returns in the first place as a courtesy to customers, because it's something people expect from large stores.
This is not true. Most states have laws on their books that say something to the effect of "Any buyer has the right to return an item within 3 days of purchase."
In addition, always buy with your credit card. Most credit cards have a 90 day return policy on any item you buy. They say, what they will do is refund your money if the merchant refuses to take the return back within 90 days. In actuality, they issue a chargeback to the merchant, and you get your money regardless. Of course the merchant is getting screwed, but remember folks, this is Best Buy. Fuck 'em. Until they stop using deceptive marketing practices like "FREE after rebate!". Of which, there are three rebates, and they purposefully disqualify you for some nitpicky reason like you forgot to dot an I on the rebate application, or each one of the 3 rebates requires the Original UPC code (not a copy), so you can only really get one of them.
The practices they have been engaging in for years are fraudulent and deceptive. They shouldn't be surprised to find out that some customers are manipulating these processes for their own gain (getting rebates after returning the items). The entire system would be far better if the FTC made rebates illegal. Everything would be one price, you would pay that price, and that's it. This will never happen though, because the government likes the higher sales tax they get, because customers pay the sales tax on the original price, not the price they get after rebate.
Oh good grief, there's nothing about Apache which makes it superior to IIS.(I can't believe people are still going around pandering this lie)
Good grief yourself. Let's see, the Russian mob has currently targetted IIS due to it's lack of security. They are spreading javascript into IIS servers, which installs keystroke loggers onto people's machines that use IE to connect to those servers, so they can harvest credit cards and other financial data. And you think Apache isn't superior to IIS? I've never heard of Apache servers falling victim to an exploit of this magnitude. Tens of thousands of IIS sites are currently compromised, and there's been no patch from Microsoft.
But the problems with IIS go beyond that. They are fundamental architectural problems. You see, Microsoft, in their infinite wisdom, decided to make IIS part of the kernel, in order to increase the speed of serving pages. Sure, this let it beat Apache in some benchmarks, but it also opened a can of worms, because now a compromise of IIS makes it possible to root the whole system. Apache runs at user level, so even if it's compromised, you still can't root the system. Sorta like the difference between Windows and Mac OS X. Almost any program executing on Windows can get System level privileges. A program on Mac OS X cannot, unless it uses setuid, and even then it will prompt you for an administrator password first. You don't see rootkits and worms going around nearly as much for OS X now, do you?
Look at the Netcraft numbers. There's a reason why Apache is used more often than IIS, and it isn't the price (although that's a nice benefit). It's just a superior product, all around.
But that doesn't mean that it's the only data rate supported. The product doesn't even exist yet.
That's not what I'm saying. The product obviously will support multiple data rates, because they're using 48k for their estimates, while it also supports music purchased on Sony Connnect which is 132k. Therefore, it must support at least those two, or it would be false advertising.
Amica Insurance, who I get my renter's insurance from, has an option called "Computer protection", which even covers things like dropping a laptop and cracking the LCD. My $25,000 policy is only $24.22 a month. BTW, $25,000 is to cover all of my belongings, not just computers.
I think it's a little early to proclaim that Sony will only support 48K data rates in this device.
RTFA. Sony only uses the 48k rate to come up with the magic 13,000 songs in your pocket figure. The normal rate of 132k (songs purchased on Sony Connect) will actually let you store less songs than you would be able to if you used a 20GB iPod.
Furthermore, one wonders if the magic "30 hours of battery life" claim is also derived using 48k files. Believe me, this makes a big difference, because battery life on HD based devices goes down in proportion to the size of your files. Larger filesizes means more time accessing the hard drive, therefore decreased battery life. That's probably why I only get 6-7 hours out of my 30GB iPod... most of my files are MP3s encoded with Lame's alt-preset-standard, which uses VBR from 128k up to 320k.
Okay, so it's been tough, but you've finally managed to stop yourself from taking Dell up on its kind offer to crush your iPod into a thin paste in exchange for $100 off one of its own stellar music players. Good for you. Only now you're finding yourself tempted by those new players that Sony introduced yesterday-- in particular the NW-HD1 Network Walkman. As faithful viewer Mike Scherer pointed out, MacMinute reports that the NW-HD1 (catchy name) has a 20 GB hard drive, but weighs only 4 ounces-- almost thirty percent less than a 20 GB iPod, and only about half an ounce more than a miniPod with a mere 4 GB storage capacity. Trust us, size does matter, as through-the-roof miniPod sales will attest; Dell's player is a clunky slab by comparison, and when we had the misfortune to encounter a 40 GB Nomad Zen last weekend, we mistook the thing for a brick wrapped in tin foil.
Oh, but the temptation doesn't stop at size; whereas the iPod claims 8 hours of use per battery charge, the NW-HD1 boasts 30. What's more, since a 20 GB iPod goes for $399 and Sony's minuscule new player will sell for "less than $400", pricing will likely be a dead heat. So let's recap, here; for the same price as an iPod, Sony offers a smaller and lighter player with gallons more juice per charge, the same size hard disk, and-- did we mention this?-- the ability to store 8,000 more songs. Really! See? Apple claims its 20 GB iPod will put 5,000 songs in your pocket, while Sony's press release insists that the NW-HD1 will hold "up to 13,000 four-minute songs." No wonder you're feeling tempted.
Well, it's cold shower time, kiddies. First of all, any sort of song capacity comparison is a joke, since a 20 GB hard drive is a 20 GB hard drive. Sony's drives aren't enchanted by a dusting of magical pixie dust before leaving the factory or anything. (At least, if they are, you'd expect Sony to play that up as a differentiating factor.) The difference in numbers here is that Apple bases its song count on 128 Kbps AAC files, while Sony's tally assumes "songs recorded at 48 kilobits per second." Yes, 48 Kbps. Considering how many people whine that even 128 Kbps AAC files don't sound good enough, we're going to go out on a limb and assume that 48 Kbps songs in any format are probably going to sound like a portable handheld AM radio playing from the bottom of a well while a few dozen people pop bubble wrap nearby.
And here's the real deal-breaker: about that format? Turns out that Sony's decided to go with its proprietary ATRAC3 format... and nothing else. While Apple pushes AAC pretty heavily (it's the only thing it sells at the iTunes Music Store), at least the iPod can also play AIFF files, WAVs, the new Apple Lossless format, and probably most importantly of all, good ol' MP3s. If you get an NW-HD1, though, you'll have to transcode your entire music library into ATRAC3 before you can carry it around with you, and believe us when we tell you that you're not going to want to do that.
See, aside from the time you'd have to invest, there's the little matter of the fact that, quality-wise, the ATRAC3 format apparently sucks eggs whole through a Crazy Straw. For evidence, we point you towards the results of Roberto Amorim's l
Slashdot Mirror
How on Earth is RAID 5 "less real" than RAID 1 or 10 ?
I mean only that using RAID 5 your write performance always suffers due to parity calculations that have to be done on the data. What suffers even worse is read-modify-write, whenever existing data needs to be modified. Your RAID controller or kernel has to read the data into memory, modify it, then calculate parity on the entire modified block, then write back out to disk. Even if you only make a 1 byte modification to a file, the read-write-modify cycle takes a huge performance hit and parity must be calculated for the entire file.
For this reason, most DBAs don't consider RAID 5 useful for serious OLTP type databases.
Yeah--oddly though I'm getting better Bonnie results by using Linux RAID 5 than their hardware RAID 5. But it is possible (just) to stream full-size, full-framerate PAL video to 'em over NFS! (sustained 40MB/s). Anyway in software you can now do RAID 6 :)
The reason why you're getting better RAID 5 results from software RAID vs. hardware RAID is because of the parity calculations involved with writing to a RAID 5 volume. On a hardware RAID setup, these are calculated on the RAID card itself, which probably has a 200 or 400 mhz. chip that does these calculations. Back when CPUs were only 400 mhz, this was great, because there was no load put on the CPU, and the RAID controller worked just as fast or faster than a software RAID setup. Now that CPUs are 3 ghz. +, there's no way a dedicated hardware RAID card can keep up, and unless you're running a huge load on the server, youv'e probably got 1 ghz. or so of free CPU bandwidth to burn for software RAID...
Want to see the performance really increase? Give up RAID 5 and go with a real RAID solution like RAID 1 or RAID 1+0.
Did companies using PowerPC based IBM servers manage to stay in business during the dot.bomb? Maybe the saying "nobody ever got fired for buying IBM" has some solid reasoning behind it.
I don't think your choice of Unix vendor can save your company if you have a failing business model. The companies that made it, like eBay, have a pretty healthy mix of Sun, HP, and probably Linux as well.
I'm curious, why do linux distributions typically have PowerPC releases always up to a generation behind when it would seem that HPPA and SPARC are likely a smaller installed base?
Is it that there are relatively few of the PPC base demand a Linux distribution when compared to what are mostly server-class CPU's and more likely to be running a Linux distro?
Well, I believe this is due to two things. First, you have the eBay phenomenon. Tons of Sun and HP hardware available for dirt cheap on eBay, and we're talking server class machines for a fraction of their dot.bomb retail prices. That, plus a lot of techs got "free" boxes when their dot.bombs went under and they just sort of "acquired" boxes that would otherwise be repo'd by the creditors.
I'm curious to know if this Sparc release of Gentoo actually frickin' works or not! I tried to install Gentoo 1.4 on a Sparc and it was a frickin nightmare. I did finally get it working (somewhat), but X barely worked, and most packages wouldn't emerge properly. I couldn't even think of getting KDE working on the thing as the emerge would just crap out on me. It seems that the Sparc maintainers don't really support or even maintain the product.
It's also not a very good sign when you go to the release website and there are no release notes available for Sparc or HPPA platforms... Gee, if they're missing the release notes, I wonder if they ever even released the software.
As usual, I'm betting this is just a "vapor" release on all platforms except for X86 and AMD64. Why does such a cool distro have to torment me so? If you can't release a decent Sparc distro, don't even waste my time announcing it. I'll just stick with Debian, which as far as I know is the only current distro that actually works on Sparc.
Try to find a host OS with a TCP/IP stack that can properly utilize 1 gigabit ethernet, let alone 10 gigabits. Hint: It ain't Linux...
Try to find a storage solution that can read or write that fast. I'm thinking something like EMC with about 6-8 2 gigabit HBAs using Veritas DMP (dynamic multi-pathing).
Try to get all of the above, along with a 133 mhz. 64-bit PCI-X bus that still can't actually keep up with 10 gigabits of data. (133 mhz. 64-bit PCI-X is only about 1024 megabytes per second, not counting overhead).
The problem is, right now, the rest of the parts of a system just can't keep up with 10 gigabit ethernet. The only box that I would use that can handle that many I/O paths to storage (we're talking six to eight 64-bit 66 mhz. 2 gigabit FC host adapters) is a Sun Fire 6800 or something larger. The problem is, Sun doesn't yet support PCI-X, so now your 10 gig ethernet card is going to be limited to a 66 mhz. 64-bit PCI version, which will only transfer a maximum of 512 MB per second, not counting overhead. That is less than half of the available bandwidth of 10 Gig Ethernet.
You can forget about putting it in any Intel based system. There are not enough I/O busses and I/O controllers in even the beefiest Xeons or Opterons that can handle this much bandwidth (to disk).
Also, if your application doesn't need to write all of that data to disk, then how large is this dataset in memory that needs to be transferred at 10 gigabit speeds? If you had a server with 64 GB of memory, it could transfer it's entire memory set over 10 gigabit ethernet in less than 60 seconds.
A far better, and more economical solution, if you really need 10 gigabits of data throughput to the network, would be to use the same Sun server, and a product called Sun Trunking, which allows you to bond multiple gigabit ethernet interfaces together. You get all of the throughput you want, plus more fault tolerance. I've set it up before, and you can have a continuous ping going, across 4 connections, and pull 3 of those 4 connections and the ping keeps going, without even a dropped packet. It's really fault tolerant, and uses your existing switches, NICs, and hardware, without forcing you to upgrade your entire core switch architecture.
Doesn't anyone else besides me see a conspiracy
theory in the making?
You're not alone at all. I got modded as flamebait for stating this in someone's journal. The DHS declaration last week that "polling places will be likely targets of terrorist attack" seems designed to instill fear in Joe Sixpack and keep him from voting our incumbent Republican government out of office.
The simple truth of the matter is that polling places are usually churches, schools, people's garages and homes. To think that somehow Al Qaeda is going to target the tens of thousands of small neighborhood polling places across the country is ludicrous.
The government that is in power will try whatever dirty tricks are necessary to win. They're not above it.
You know what is really amazing... The mayans had a very sophisticated calendar which was entirely based around the cycles of the Sun's magnetic fields. I heard an interesting speech by Maurice Cotterell, an engineer who heavily researched the Mayan calendar, and he predicts that in the year 2012, there will be some type of major magnetic field shift in the magnetic fields of the earth at this time. He doesn't believe the world will end then, but he believes that there could be major catastrophes because of this. If you get a chance, I recommend reading one of his books and deciding for yourself whether what he says is believable.
Maybe I can use this to track down the scumbags who send junk faxes to me at all hours of the night and morning, but whose numbers are listed only as "Out of Area". In fact, I bet this would be a handy tool for those who are trying to stop these asshats.
It won't be too long before these asshats figure out how to spoof on their own, then it will be just as bad as email spam where the headers give you no clue as to the actual source of the email (other than the sending MTA's IP address).
Kinda hard to surf the net at the speeds mentioned in the article when carriers like Sprint haven't made 3G fully functional. I am still only able browse the web at a mere 5k or so, 10k bursts if I'm lucky.
Really? I consistently get 160kbps with my Sprint Treo 300 connected through a custom PDANet driver to my laptop (Windows only, sadly). There is also a program called WirelessModem for the Mac, but it's so buggy it caused kernel panics every few hours or so (grey screen of death) on my PowerBook.
Sprint uses 1xRTT for data, who's theoretical maximum is (I believe) 224kbps. Of course, as with any wireless protocol, there is overhead. The reason you're seeing slower speeds is (I'm just guessing here) probably because you're connecting over Bluetooth. Although the Bluetooth spec supports speeds up to 768kbps, most BT phones serial port profile only supports 115200 bps (just like a regular serial port), so you're limited by the bandwidth of that connection. With a Sony Ericsson T610 I can only get about 5KB (kilobytes) per second, even when doing a Bluetooth file transfer, so it's pretty unlikely that I'd get any better than that using the Internet.
So, in conclusion, if you want 160kbps through your 1xRTT Sprint PCS connection, get a phone that supports a USB connection instead of Bluetooth. The difference is night and day.
You're a wonderful example of the ignorance which is so pervasive through this industry.
.NET (which isn't that bad in itself, except for the fact that MS seems to change dev tools every month now so if you want to keep porting your code, go right ahead), using a webserver (IIS) that's riddled with holes.
Nice... I just checked out your personal website, and I see you've bought the MS party line hook and sinker. Let's see, you're using a notoriously riddled with security holes CMS (Nuke), running on top of
You still haven't rebutted my initial argument that user-level webservers will always be more secure than kernel-level webservers. Care to try?
Microsoft provided a patch for this IIS exploit back in April.
I should have clarified that statement. There's currently no client patch available for the millions of IE users that are most at risk of having keystroke loggers installed.
You're a wonderful example of the ignorance which is so pervasive through this industry.
And you sir, are a Microsoft shill. No doubt the majority of your income comes from foolish customers that are simply too ignorant to know that there are superior free solutions to the expensive MS solutions you recommend on a daily basis. Besides, all the support calls sure boost your billable hours. If I were in need of a corporate website, your company would be the last one I'd contact, for that very reason.
What if you bought it with your Best Buy card?
That would leave you without the option to dispute it wouldn't it?
Probably. I don't like store cards for that, among other reasons. The main one is that they charge some ungodly interest rate like 22%.
Cisco Press books are, without a duobt, the best technical manuals (from a manufacturer) that I have yet read.
I agree wholeheartedly. Especially Basam Hallabi's Internet Routing Architectures. (No affiliate link) This book taught me how to establish BGP routing policies, and is considered fundamental reading by almost anyone on NANOG.
It's good to know how the OSI model works, especially when troubleshooting tricky layer 2, 3, and 4 network issues. That being said, when I took the CCNA test they only had about 2 questions on it. They were something like "Which of the following are layers in the OSI model" or something really basic like that. Anyway, my instructor had a really good way to remember the OSI model that I still remember to this day (took the class in 1997):
All People Seem To Need Data Processing.
The letters in this saying correspond very nicely with the layers in the OSI model:
Application
Presentation
Session
Transport
Network
Data Link
Physical
Hope this helps those of you studying for your CCNAs (judging by enrollment numbers in these college courses, there's a lot of you.)
If you register at Lulu, the free download is a PDF.
Wow, Lulu.com is Slashdotted already... There must be a pretty huge demand for this book. I know I haven't been CCNA certified since version 1 of the test (they are up to 3 now), but I'm downloading the Word version because someday I'd like to update that certification.
... It's things like this that make me want to go back to the store. Of course when you talk rebates though, CompUSA is definitely the KING of rebates.
Are you and I thinking of the same store? CompUSA is the worst at returns. They have a 15% restocking fee on everything, even if it's in brand new condition in the original packaging. Let me tell you my CompUSA story:
A couple years ago, I was in the market for a PDA. I'd heard good things about the Handspring models, so I went in and purchased one (can't remember which one) for $300. It was pretty nice, but the next day I was reading Slashdot and saw an article on the Sharp Zaurus SL-5500. It ran Linux, had a ton of open source software available for it, and was only $100 more expensive than the Handspring. I had to have one. So, I immediately wiped the memory on my Handspring (put it back to factory defaults), boxed it back up in the original packaging, and took it to CompUSA to exchange it for a Zaurus. It was in brand new condition, barely used, no scratches, with the original receipt. I wanted to exchange it for the Zaurus, and I was willing to pay the $100 difference between the two models.
The manager tells me that he is going to charge me a 15% restocking fee on the Handspring to return it, even though I was buying a more expensive model at the same time! Fucking ridiculous. I sat there and argued with him for about 30 minutes. I showed him receipts proving that I had spent $2000 there in the last month, on everything from printers to computer accessories to software, and none of it returned. Not only that, I was paying full retail price for the Zaurus just because I wanted it right now. He wouldn't budge. I told him flat out that I would never shop at CompUSA again, and that he had lost a valuable customer. I returned the damn Handspring, -15% of course, and I've never shopped there since.
CompUSA doesn't deserve any of my business. They also have some of the worst prices in the industry. You can occasionally get a good deal there if you look for the rebates or loss leader items, but it's not worth the 15% restocking fee if you have any problems with it. Now I buy all of my computer hardware from NewEgg, and I've never looked back. Newegg is mail order, but they've got some of the best customer service I've ever seen. I had a memory DIMM fail once after 8 months of use, and all I had to do was request an RMA from their website, ship it to them, and the very next day they sent me a brand new replacement, which arrived in 2 days from FedEx. All in all I was only without the DIMM for a week. Same thing happened with a video card that stopped working after a few months. I got a brand new one which happened to be better than the one I returned, within a week. I guarantee you that even with the PSP or whatever purchase protection plan they have, you're going to wait a couple of weeks minimum to get your repaired merchandise back from CompUSA.
Never again will I shop at CompUSA. Best Buy is just as bad. If I had to pick one retailer to get "instant gratification" shopping, where you want the item right that second, it would be Circuit City, just because they aren't quite so draconian about returns and exchanges. They even have that 110% pricematch thing, so if you see it for cheaper next week in a Best Buy ad you end up getting a sweet deal.
Under this plan, I tried taking the stereo back because it wasn't functioning properly.
The moral of this story: Always purchase electronics with a credit card. If the merchant won't take it back and immediately exchange it for a working version of the same product (no 2 weeks waiting for repair, I'm talking about immediate exchange right there in the store), call your credit card company and dispute the charge. Works every time. The fact of the matter is that the credit laws protect you because you do not have to pay for goods or services that you never received, or were defective. The credit card company can't bill you for them and will issue a chargeback which means the merchant won't get their money either.
Also, if a merchant gets too many chargebacks in a certain period of time, the credit card company will revoke their merchant account, leaving them unable to accept credit cards period.
This is also a highly effective bargaining tool if the customer service rep is giving you a hassle about returning the item. Just tell them you will dispute the charge with your CC company if they don't accept the return right this second. They will immediately cave, because they know they don't have a chance of winning that battle.
A business is under no obligation legally to take back a returned item once sold, as long as it is not defective and was not sold under false pretenses. Businesses like BestBuy take returns in the first place as a courtesy to customers, because it's something people expect from large stores.
This is not true. Most states have laws on their books that say something to the effect of "Any buyer has the right to return an item within 3 days of purchase."
In addition, always buy with your credit card. Most credit cards have a 90 day return policy on any item you buy. They say, what they will do is refund your money if the merchant refuses to take the return back within 90 days. In actuality, they issue a chargeback to the merchant, and you get your money regardless. Of course the merchant is getting screwed, but remember folks, this is Best Buy. Fuck 'em. Until they stop using deceptive marketing practices like "FREE after rebate!". Of which, there are three rebates, and they purposefully disqualify you for some nitpicky reason like you forgot to dot an I on the rebate application, or each one of the 3 rebates requires the Original UPC code (not a copy), so you can only really get one of them.
The practices they have been engaging in for years are fraudulent and deceptive. They shouldn't be surprised to find out that some customers are manipulating these processes for their own gain (getting rebates after returning the items). The entire system would be far better if the FTC made rebates illegal. Everything would be one price, you would pay that price, and that's it. This will never happen though, because the government likes the higher sales tax they get, because customers pay the sales tax on the original price, not the price they get after rebate.
Oh good grief, there's nothing about Apache which makes it superior to IIS.(I can't believe people are still going around pandering this lie)
Good grief yourself. Let's see, the Russian mob has currently targetted IIS due to it's lack of security. They are spreading javascript into IIS servers, which installs keystroke loggers onto people's machines that use IE to connect to those servers, so they can harvest credit cards and other financial data. And you think Apache isn't superior to IIS? I've never heard of Apache servers falling victim to an exploit of this magnitude. Tens of thousands of IIS sites are currently compromised, and there's been no patch from Microsoft.
But the problems with IIS go beyond that. They are fundamental architectural problems. You see, Microsoft, in their infinite wisdom, decided to make IIS part of the kernel, in order to increase the speed of serving pages. Sure, this let it beat Apache in some benchmarks, but it also opened a can of worms, because now a compromise of IIS makes it possible to root the whole system. Apache runs at user level, so even if it's compromised, you still can't root the system. Sorta like the difference between Windows and Mac OS X. Almost any program executing on Windows can get System level privileges. A program on Mac OS X cannot, unless it uses setuid, and even then it will prompt you for an administrator password first. You don't see rootkits and worms going around nearly as much for OS X now, do you?
Look at the Netcraft numbers. There's a reason why Apache is used more often than IIS, and it isn't the price (although that's a nice benefit). It's just a superior product, all around.
Gee, I guess the Always On network isn't really always on...
But that doesn't mean that it's the only data rate supported. The product doesn't even exist yet.
That's not what I'm saying. The product obviously will support multiple data rates, because they're using 48k for their estimates, while it also supports music purchased on Sony Connnect which is 132k. Therefore, it must support at least those two, or it would be false advertising.
Amica Insurance, who I get my renter's insurance from, has an option called "Computer protection", which even covers things like dropping a laptop and cracking the LCD. My $25,000 policy is only $24.22 a month. BTW, $25,000 is to cover all of my belongings, not just computers.
Recursion alert - so are you!
I think it's a little early to proclaim that Sony will only support 48K data rates in this device.
RTFA. Sony only uses the 48k rate to come up with the magic 13,000 songs in your pocket figure. The normal rate of 132k (songs purchased on Sony Connect) will actually let you store less songs than you would be able to if you used a 20GB iPod.
Furthermore, one wonders if the magic "30 hours of battery life" claim is also derived using 48k files. Believe me, this makes a big difference, because battery life on HD based devices goes down in proportion to the size of your files. Larger filesizes means more time accessing the hard drive, therefore decreased battery life. That's probably why I only get 6-7 hours out of my 30GB iPod... most of my files are MP3s encoded with Lame's alt-preset-standard, which uses VBR from 128k up to 320k.
Okay, so it's been tough, but you've finally managed to stop yourself from taking Dell up on its kind offer to crush your iPod into a thin paste in exchange for $100 off one of its own stellar music players. Good for you. Only now you're finding yourself tempted by those new players that Sony introduced yesterday-- in particular the NW-HD1 Network Walkman. As faithful viewer Mike Scherer pointed out, MacMinute reports that the NW-HD1 (catchy name) has a 20 GB hard drive, but weighs only 4 ounces-- almost thirty percent less than a 20 GB iPod, and only about half an ounce more than a miniPod with a mere 4 GB storage capacity. Trust us, size does matter, as through-the-roof miniPod sales will attest; Dell's player is a clunky slab by comparison, and when we had the misfortune to encounter a 40 GB Nomad Zen last weekend, we mistook the thing for a brick wrapped in tin foil.
Oh, but the temptation doesn't stop at size; whereas the iPod claims 8 hours of use per battery charge, the NW-HD1 boasts 30. What's more, since a 20 GB iPod goes for $399 and Sony's minuscule new player will sell for "less than $400", pricing will likely be a dead heat. So let's recap, here; for the same price as an iPod, Sony offers a smaller and lighter player with gallons more juice per charge, the same size hard disk, and-- did we mention this?-- the ability to store 8,000 more songs. Really! See? Apple claims its 20 GB iPod will put 5,000 songs in your pocket, while Sony's press release insists that the NW-HD1 will hold "up to 13,000 four-minute songs." No wonder you're feeling tempted.
Well, it's cold shower time, kiddies. First of all, any sort of song capacity comparison is a joke, since a 20 GB hard drive is a 20 GB hard drive. Sony's drives aren't enchanted by a dusting of magical pixie dust before leaving the factory or anything. (At least, if they are, you'd expect Sony to play that up as a differentiating factor.) The difference in numbers here is that Apple bases its song count on 128 Kbps AAC files, while Sony's tally assumes "songs recorded at 48 kilobits per second." Yes, 48 Kbps. Considering how many people whine that even 128 Kbps AAC files don't sound good enough, we're going to go out on a limb and assume that 48 Kbps songs in any format are probably going to sound like a portable handheld AM radio playing from the bottom of a well while a few dozen people pop bubble wrap nearby.
And here's the real deal-breaker: about that format? Turns out that Sony's decided to go with its proprietary ATRAC3 format... and nothing else. While Apple pushes AAC pretty heavily (it's the only thing it sells at the iTunes Music Store), at least the iPod can also play AIFF files, WAVs, the new Apple Lossless format, and probably most importantly of all, good ol' MP3s. If you get an NW-HD1, though, you'll have to transcode your entire music library into ATRAC3 before you can carry it around with you, and believe us when we tell you that you're not going to want to do that.
See, aside from the time you'd have to invest, there's the little matter of the fact that, quality-wise, the ATRAC3 format apparently sucks eggs whole through a Crazy Straw. For evidence, we point you towards the results of Roberto Amorim's l