Drive manufacturers take a new hard drive, run a hundred drives or so for some number of weeks, and measure the failure rate. Then they extrapolate that failure rate out to thousands of hours... So, let's say one in 100 drives fail in a 1000-hour test (just under six weeks). MTBF = 100,000 hours, or 11.4 years!
To make this sort of test work, it must be run over a much longer period of time. But in the process of designing, building, testing and refining disk drive hardware and firmware (software), there isn't that much extra time to test drive failure rates. Want to wait an extra 9 months before releasing that new drive, to get accurate MTBF numbers? Didn't think so. How many different disk controllers do they use in the MTBF tests, to approximate different real-world behaviors? Probably not that many.
Could they run longer tests, and revise MTBF numbers after the initial release of a drive? Sure, and many of them do, but that revised MTBF would almost always be lower, making it harder to sell the drives. On the other hand, newer drives are certainly available every quarter, so it may not be a bad idea to lower the apparent value of older drive models.
So, it's better to assume a drive will fail before you're done using it. They're mechanical devices with high-speed moving parts, very narrow tolerable ranges of operation (that drive head has to be far enough away from the platters not to hit them, but close enough to read smaller and smaller areas of data). Anyone who's worked in a data center, or even a small server room, knows that drives fail. When I've had around two hundred drives, of varying ages, sizes and manufacturers, in a data center, I observed a failure rate of five to ten drives per year. This is well below the MTBF for enterprise disk array drives (SCSI, FC, SAS, whatever), but drives fail. That's why we have RAID. Storage Review has a good overview of how to interpret MTBF values from drive manufactures.
I think Chess is right - the Cone of Silence originated with Get Smart and was completely useless. I mean, it's a joke, and Mission: Impossible wasn't (intentionally) funny, though they did use the idea. But calling it a staple is an overstatement - laugh tracks, harried husbands, adorable children and perfect housewives were staples of 60's TV. Not like our modern, thought-provoking entertainment products...
Sure, it will still be a state-issued driver's license (or non-license ID). And many states already meet the pre-license verification and anti-counterfeiting measures required by RealID. But the bigger problem is that it all goes to a "secure," consolidated Federal database. What are the safeguards against abuse, tampering, or theft? DHS won't say. How are they maintaining this massive data warehouse? Nobody knows.
Some people don't value a sense of humor but, for me, it's important to know that the people I hire not only can do the job, but they're someone I want to hang out with 40-50 hours a week..
Careful on that slope buddy. I've been there, it's slippery. As long as all the people you interview are the same, in terms of discrimination, you're in good shape. But if some are older, or different genders, or different religions, or different ethnicities, you could set yourself and your company up for an employment discrimination lawsuit.
"But no," you say, "I'm just filtering for best fit." And that might work, more or less, but you pretty quickly come across people who are different from you, and if you turn them away because, let's say, they don't get your jokes, or they're not funny, and they happen to be a woman/Muslim/Asian/person-over-50... Well, what's the difference between discriminating based on personality (highly subjective) and discriminating based on race (also highly subjective)?
Now, age/sex/religious discrimination lawsuits don't happen that often in hiring situations, but if you're considering candidates for internal promotion, it gets pretty sticky.
Of course, fundamentally, the problem I see with your approach is that it boils down to hiring people you like. And in that pool of interviewees, people you've just met, the ones you like are the ones most like you. So you end up hiring people like you, turning away people who are different, and this hurts your organization. You need diversity of opinions, ideas, and personal experience to solve problems. Does your company want you to be comfortable, or productive? OK, team cohesion is important, but only because it helps performance - it's a means, not an end. Sure, you can get by with an army of clones, but you'll be doing yourself and your company a disservice.
The C:\ prompt, first used in CP/M, is at least 30 years old.
But look at all innovative, closed-source operating system software running the world now!
VMWare ESX - based on Linux kernel with incremental enhancements
Windows NT/2000/XP/2003/2008 - designed by DEC VMS architects to use a microkernel architecture. Remind you of anything? Something that rhymes with "tunics"?
MS DOS - closed source software which appropriated most of its design from CP/M...
IBM OS/2 - DOS with a GUI and protected-mode memory access
IBM System z9 - UNIX, based on zSeries, based on System/390, based on System/370, based on System/360...
This is one of the worst networking articles I've ever read. There is nothing remotely interesting or informative in those two pages. There's not enough detail to offer any insight into networks, project management, system design or anything else. Why?
Some phone companies have figured out that the can actually make more money (sell more circuits) by lowering the price without increasing their costs all that much.
That's oversimplifying the service offerings and associated costs.
Verizon didn't lower their service charges by ~80%, undercutting their profitable T1 sales to business and institutional customers. They can only make more money by adding customers. FIOS primarily competes with cable-modem ISP services. It's a good deal compared to a cable modem and for a business which can't justify the cost of a T1, it looks like a good service.
But FIOS for business does not provide an SLA equivalent to Internet T1 service.
Speed and uninterrupted use of the service are not guaranteed.
This is one of the main reasons why FIOS for business starts at $69/month and a Verizon (or similar) T1 with Internet service starts at around $600/month. The upload speed, average latency, and guaranteed latency are important factors as well.
In terms of variable costs, the SLA is probably the most expensive part of the service in terms of labor and management/monitoring infrastructure. Verizon has plenty of infrastructure to manage and monitor T1's, ISDN lines, POTS lines and other high-volume telecom circuit types. But FIOS uses different distribution, cable plant, and customer premises equipment than all the other Verizon services. This creates an entire new set of costs. This is why they don't offer it for FIOS - it's too expensive to build out the management systems, add NOC staff, and build the new FIOS network services at the same time.
FIOS may be offered with more substantial service guarantees in the future - I hope it is - but right now Verizon is learning how to operate their FIOS networks with consumers and small businesses as guinea pigs.
Sure, some are paid more than their productivity or merit might dictate, but the vast majority are underpaid, overworked, and without sufficient resources. I have yet to meet a teacher who doesn't purchase materials for their classroom (forget reimbursement). Nobody congratulates a new teacher on their personal-finance acumen. Communities rarely have enough money to maintain manageable class sizes (i.e. 20 or fewer students per teacher). But our political leaders and experts talk endlessly about how important children and education are - our future, etc. The teachers unions have a point - once you start differentiating between academic subjects, you create subjective differences all over the place. School system budgeting will become even more difficult to understand and time consuming than it is. You can create endless rules to encourage better teaching, but that won't improve the pool of applicants for teaching jobs.
So, double their pay. Across the board. Not just math and science teachers, but everyone.
"Oh no!" you say, "then we'll just be paying the bad teachers and the good teachers!" Well, true, but so what? Teachers are retiring at record rates. Perhaps you've heard of this "baby boomer" generation? Qualified, talented young teachers leave teaching because they don't get the pay/support/supplies they need. We need more teachers, and we need better teachers. Supply, meet demand.
Blizzard is already doing this. My company uses colo space in two facilities that also house Blizzard servers. I don't know the voltage, but all of their World of Warcraft server farms (~1000 blades each, around a dozen sites) run HP blades on DC power. The data centers have serious AC power distribution systems, but they've built custom DC power distribution in their floor space. The grounding wires - they're not small.
Cooling is provided by the facility's HVAC system, it's paid for (it's hard to bill each customer for hvac when they're all in a room the size of a football field). So Blizzard seems to be using DC to reduce total power consumption. The other benefit, as the parent suggested, is reduced wear on servers when they're not dealing with heat produced by inefficient AC/DC conversion. Of course, high-end power supplies are getting more and more efficient, but there's always some power loss/heat generation.
Even though exploits and bugs will continue to surface long after MS stops releasing patches, you'll save yourself a lot of risk if you log in as an unprivileged user. If you use the NTFS file system (and fixacls.exe if you have to convert after the install) your general-purpose login will have very few ways to wreck the system. Sure, there are privilege escalation attacks, but you'll be protected from many common bugs and exploits.
Keep a local admin account, or two, using a strong password. Change the default Administrator to a different user name and only use the admin priv's for maintenance.
I'm skeptical of any company's "study" of its own product and this one has several flaws. The authors compare Windows 2000 Server and Windows Server 2003 with "Fedora Core" but never mention which version of Fedora they're using. Fedora Core 5 and 6 came out this year - are they using one of these disributions or something older?
Despite the study's flaws, many of the posts in this thread seem to focus on the inherent, perhaps necessary, superiority of Linux sysadmins to Windows sysadmins. This tangential argument actually makes the Microsoft argument. In fact, this has been the essential MSFT sales pitch for about 20 years - "our products are easier to use" - regardless of the competition. Sure, they sometimes try "our products are better", but the ease-of-use argument is their bread and butter.
The perceived, and fairly real, problem with Linux administration is that you have to do some level of programming. You don't have a choice. Most open-source (and commercial!) Linux applications require some level of configuration by hand, most commercial (and open source!) Windows applications do not. Since that manual configuration takes time, I suspect it diminishes the overall productivity of the skilled Linux admin (with requisite programming skills).
Can a Linux admin/developer really manage more servers than a Windows admin? My experience is that they can't - it tends to balance out. A well-configured, properly sized Windows server and a well-configured Linux server will both tend to run for quite a while without much intervention. Sure, many Linux applications are more memory and processor efficient than their Windows competition, and for large-scale environments this may be a decisive factor. But for most environments, even with dozens or hundreds of servers, the flexibility of a system in which you can modify anything, down to the kernel, doesn't provide enough business value to offset the high salary costs for sysadmins-with-programming-skills compared to those without. For people who don't know how to write shell scripts, or perl, or PHP - i.e. most of the world - Linux looks harder to use because it *is* harder to use.
Don't get me wrong - the open source approach is much more elegant - I'm just not sure it provides enough cost justification for most businesses. By arguing that Linux admins are required to be smarter and, incidentally, better paid than Windows admins, you're arguing against Linux adoption. Keep making this argument - Microsoft PR thanks you.
How often have you heard that the new version of Windows is "more secure" than the last version? A quick recap:
Windows 3.1 - no real security, but it's prettier than DOS!
Windows for Workgroups 3.1.1 - now with a login screen (but still no real security)!
Windows NT 3.51 - now with ACL's (and mostly not compatible with Win3.1 apps)!
Windows 95 - also has a login screen! no real security, but prettier than WfW!
Windows NT 4.0 - now with shared ACL's (domains) - the most secure Windows ever!
Windows 98 - Slightly less likely to crash than Win95! No NT security features!
Windows ME - Now with some system-software protection, but still no ACL's!
Windows 2000 - An improved interface and kernel! Active Directory 1.0! Now, the most secure Windows ever!
Windows XP - The successor to the Win2k and Win9x kernel products - super duper secure! Home users still run as the super-user, but it's less likely to crash! ACL's for Professional users and a very limited firewall make this, yes, the most secure Windows ever!
Windows 2003 (server) - The XP kernel in a server! Hardly anything runs by default! The Most Secure Windows Ever!
Windows Vista - Still with ACLs! New ways to limit access! Everyone's running as superuser, but with more warnings!
Windows Longhorn (server) - Not fully designed, but looks a little less secure than Win2003 - possibly *not* the most secure Windows ever!
So, this is an opinion piece, tying together loosely related facts, but there isn't much to worry about. Look at the product lines at these two companies - they make all kinds of non-AV software. These large ISV's get more insight into MSFT product road-maps than just about anybody - they knew this was coming.
I'm not sure where all the hatred for AV vendors comes from. Sure, you can't run a Windows system without an AV package, but that's not the fault of the AV vendors. These guys have been filling a need for DOS and Windows users for about 20 years. You don't have to buy their software, but their existence isn't hurting anyone. There are plenty of free and smaller-circulation commercial AV packages out there if you don't like these two. Use them instead. But to say "these companies must die" - really, why again? Because their software doesn't always work? Isn't that true of all software? Because once an AV package "messed up your computer" and you had to reinstall Windows? Maybe. More likely, your computer was already broken and an AV software installation was the straw that broke the camel's back.
AV vendors deal with a lot of confused end-users because their apps get deep into the OS and can get stomped by bad device drivers, OS changes, and other 3rd party software. AV software is often the canary in the coal-mine, alerting you to some underlying problem (your video driver has a memory leak, somebody's installer rolled-back an OS file to a previous version, etc).
If you don't like AV software, don't run Windows. Oh, wait, you don't? So what is your point exactly?
And on the prevelance of MS Access - that prevelance is from home-grown database applications, not so much 3rd party vendors who can include MSDE with their products at almost no cost (with a Dev software license). If you're going to hack together your own Access database, it doesn't really matter what you use for a file server. With a little practice, any decent Access db programmer could build the same thing in MySQL (and get a free db server in the deal, if you want) or OpenOffice Base.
FTA: Well, let's take World of Warcraft as an example. Let's say there are around 6 million subscribers for the game. I'd say that 40 percent of the players are addicted.
Let's say there are around two million regular Slashdot readers. I'd say 45% of Slashdot readers are puppies. Cute, cuddly, Linux-friendly puppies.
So for the average user, being able to open the box and not plug in any wires (except maybe power) is a god send.
But that's not what 99% of wireless devices do - they need some wires configured before you can do anything. Wireless networking needs a wired access point, wireless mouse and keyboard need USB or PS2 wires to the box, and so on.
There are some nice opportunities for infrared and bluetooth devices to communicate with PC's, but this is an unplug-and-pray operation. Some things work together, some things don't.
As far as the assertion that Bluetooth rises up to allow all of our devices to sync with one another and the operating system , allow is the key word in that sentence. The only painless bluetooth operation I've seen is between two cell phones from the same manufacturer. They could share contacts very nicely. But connecting different bluetooth devices, or a bluetooth-enabled device and a Windows or Linux box, is not painless or predictable. Mac's seem to do this a little bit better than Windows/Linux (no big surprise there).
I have two "media pc's" in my living room - they're both Shuttle SFF pc's running Windows and Linux. No TV. No VCR. DVD's play just fine on either PC, with better picture quality than the TV I gave away two years ago. I don't use a remote, but I don't have a stack of tv/vcr/dvd/stereo remotes to contend with, so it's not a big loss. I'm not running MythTV or a Windows Media Center OS, but I play DVD's and video from iTunes and other sources on the computers. These devices have replaced a TV and VCR in my home. If I wanted TV reception, I could add TV tuner cards and get an antenna or a cable TV subscription, but I'm not really interested in TV shows
Lots of people use Mac Mini's for this purpose - they're just not called "Media PC's" - but they're small and quiet. The new iMacs come with a remote control and a quality LCD - I'd happily watch movies on those boxes.
The problem with the media PC, the convergence device, and so on, is that it isn't one pretty, easy-to-track product like an iPod or a Walkman. It's MythTV or TiVo (a little stretch, but not a big one) or a generic Linux/Windows/Mac computer which is, very slowly, replacing TV as media-consumption device. The fact that the PC architecture is rapidly evolving and quite flexible makes it attractive to people with only "hobbyist" level technical skills and circumvents the marketing efforts of PC and consumer-electronics manufacturers. The companies selling Home Entertainment PC's haven't found the right mix for a huge number of people to embrace, so people roll their own. Sure, most people over 40 would rather have super-simple consumer-electronics devices which "just work" - good for them! People in my age group, 30-40, may be a little more experimental, and I think people under 30 who are comfortable with computers and the Internet have even less use for TV.
So I think the Media PC is here, and getting bigger, but it will be hard to chronicle accurately for a few more years while the hardware/software/consumer-electronics vendors figure out what people really want and how to sell it to them.
This happens all the time, with all antivirus software. Signature-based or heuristic or both, they make mistakes. This sounds like a small mistake - it's much worse when you trigger off a widely used application or OS file.
Sure, this is one of the dangers of any malware detection software, but how else would you do it? You need to identify dangerous software using the smallest possible identifier or behavior. Sure, you could have a whole virus-library of malicious code in your detection system, but you would use a ton of memory and disk space to do it and the product would be unusable.
Same for heuristics - you want to exercise, in a pseudo-processing environment, every function in every piece of software on your machine to make sure it doesn't try anything nasty? You would lose most of your cpu cycles to heuristic analysis, waiting several minutes for a program to start executing *your* instructions after the malware analysis completes. Good luck with that project.
With 95% of the world's computers running Windows, everybody needs antivirus software. Unfortunate, and largely MSFT's fault, but there we are. NAV, SAV, McAfee, Grisoft, Trend, Sophos, etc. - they all have their pro's and con's.
Sure, it would be great to have an efficient anti-virus system that didn't depend on too-simple heuristics or too-hard-to-maintain signatures. Does someone know how to do that? There's a big AV market out there. Go ahead, build it.
This is the whole problem with integrating consumer entertainment media (CD's, DVD's, etc.) and general-purpose computing devices. General-purpose hardware with general-purpose operating systems are good at lots and lots of things. That's why people buy billions of $US worth of computers and software every year. But these general-purpose devices don't fit the expected distribution conditions for entertainment media companies - they can't handle an environment in which the consumer can change the use of the media. The DVD and CD players work just fine in this model - they just play the media.
But the machine with the keyboard is trouble because of the keyboard and the general-purpose OS environment. Microsoft started down the road to DRM-friendly media-player-only operating systems with Windows Media Edition, but this version of Windows still maintains all the existing Windows functionality.
This could work out well if an enterprising OS vendor merged cheap PC's with a media-player operating system to create single-purpose appliances. But the general-purpose OS genie is out of the bottle and hardware isn't cheap enough yet (no $50 computers, plenty of $50 DVD players) for people to give up web-browsing and text editing and email so they can have a dedicated media-player. So these two worlds are joined and general-purpose computing is already suffering.
As long as Intel maintains this architecture, with a single data bus for RAM, PCI, PCI-e, AGP, BIOS, and other integrated functions, they'll be behind AMD. AMD's current (and future) HyperTransport provides a wider, more efficient data path than the front-side bus. AMD's per-processor memory controller scales past two sockets in a way that Intel just can't match. By pushing fully-buffered DIMM's, each with its own memory controller, Intel is ceding the design point to AMD: a single memory controller is too much of a bottleneck, the load needs to be spread around. This is especially true when you go beyond two processors in a machine, but even dual-socket boxes benefit from distributed memory controllers. Sure, the Bensley FSB goes to 1033Mhz from 800Mhz, but that doesn't sound like a big jump.
Until Intel has a real answer to HyperTransport, they'll be losing the high-performance, 4+ sockets market to AMD. For smaller two-socket servers, Intel will have to pay the RAM and/or server vendors to make FB-DIMM's price competitive with different flavors of DDR.
I do some work in financial services, including working with SWIFT.
SWIFT is a network, not a database. It can store and forward data, but only for short periods of time.
SWIFT transactions can be done directly between organizations over the network, without stopping at SWIFT systems, making them slightly harder to track.
SWIFT is run by its 7,000+ member organizations, from all over the world. Most SWIFT members are banks, and the largest banks have the most influence among the members. So it's not US-based, but it's well stocked by US banks and corporations. It's not like the United Nations - the biggest players run the operation.
Since SWIFT is not a database, and not all transactions land on SWIFT computers, you can't really "query SWIFT" about financial data. You can only "query SWIFT" about the status of a message from party A or to party B. So to do data mining on SWIFT data, you would need to build your own SWIFT data warehouse and your own (massive) matrix of data relationships.
SWIFT is a network, like the Internet, but entirely private. The problem of gathering data from SWIFT communications is similar to the challenge in gathering data from AT&T's Internet backbone routers. But filtering the Internet data is easier - you can probably drop bit-torrent traffic, ISO's, binary software packages, iTunes media, etc. That's a lot of packets to ignore. If you're building a comprehensive financial database from uncorrelated data, in a multitude of formats and languages, you have to start by recording everything just to make some sense of the data.
Also, why would terrorist groups use banks? For the convenience? Certainly, wire transfers and ATM's are the *easiest* way to move money around the world. But if you were trying not to leave tracks, you probably would avoid banks or use one-time-use bank accounts for individual money transfers. People have been solving this problem in organized crime for decades, why wouldn't the same methods be available to Bin Laden or anyone else?
I manage a few small data centers and we depend on RAID 1 and RAID 5 (and redundant servers) to keep our business running. Down-time is expensive, but rebuilding a machine from scratch is expensive too. So we don't use any software RAID or software-assisted on-board "controllers" - it's hard to call them controllers. RAID Filters? RAID bridges? RAID-like adapters? I love real RAID controllers. Everybody I interview has to explain RAID and something about how it works. If someone tells me they use RAID-0 for performance, they'll need a really impressive story about why that's not plain stupid.
People like to say "RAID-0 is great for gamers and testers - it's cheap and fast," but think about the scenarios where this applies:
1. Cheap!!
You have one computer, you use it (mostly) for gaming. Your critical data exists elsewhere (GMail? cd-rw's?). A second, identical hard drive runs you about $100 US. You have on-board RAID, which relies on your processor for all the heavy lifting, but it's paid for already. When you're loading your multi-gigabyte game, the two spindles go pretty fast together and your processor, while not idle, is mostly waiting for the disks. So you can spare the CPU cycles the RAID "controller" uses, great. But this computer is twice as likely to crash and burn as a single-HDD machine, so you can't rely on it to hold any persistent data (personal finance, schoolwork, download-only software, etc, etc). You just diminished the value of your computer, and you're trying to save money.
2. Fast, but just for games!!
You have a dedicated computer, with two-disk RAID-0, and it's your gaming rig. It's optimized for speed. The on-board RAID controller isn't fast enough when you've got to load textures from the hard drive and track hundreds of independant 3d objects in (insert blow-up-stuff game title here). So you have a hardware-based controller, with its own BIOS and OS-independant management tools. You could buy three RAID cards like this for the cost of a "mainstream" video card, the cost isn't a big deal - you can afford to spend $2k on an unstable drag-racer of a gaming rig.
Someone in the world thinks your time isn't worthless, but you're willing to burn a whole day, now and then, to rebuild your computer? Why not spend another ~$200, get two more hard drives (I know you've got the fans/power supply/drive bays/liquid-nitrogen-cooling to handle it), and run RAID 1+0? You get the same speed boost, but now you don't need ***a whole other computer*** to store your "real" data.
Don't get me wrong, big, resource-intensive games are great stuff. And faster loading is great too. But how many computers do you really want to own and maintain? I know, for many people, having six machines to manage them is cooler than one, but my time is more valuable than that (I know, we're mostly talking about Windows, and you can maintain thousands of *nix boxes with perl-scripts and ssh, thanks, got it). I'd rather just get a faster drive and controller - SATA2 is pretty damn fast for a non-SCSI system.
3. Fast for testing!! No real data!! You're not testing performance, right? Because you're not depending on RAID-0 to handle real data when the test is complete... If you have a grid computing farm, and the time to rebuild a single box is only a few minutes, and you can stand some data loss, this might be a good solution. Maybe your application can withstand individual machine failures and the raw speed is more important... In this kind of setup, you're much more likely to have network bottlenecks than disk bottlenecks -- your processor-intensive tasks are passed around the network and not too much data has to be stored on disks. But if this is the case, and you can afford a grid/server farm, I doubt you're going to solve a disk-IO bottleneck with RAID 0 instead of external storage (SAN, big SCSI array, big SATA/SAS array, etc).
There probably are situations where the additional data-loss risk is offset by performance gains, but usual pro-RAID-0 scenarios in this debate just don't make sense.
The Mozilla suite, now SeaMonkey, has had user-profile sync'ing to FTP servers for about a year. SeaMonkey 1.0.2 has roaming profile functions to HTTP or FTP servers as a native feature.
What sort of testing did you do? Was it something like "3 linux DNS servers with 10k concurrent requests per minute for 60 minutes on different platforms"? Or more like "I loaded Ubuntu on ESX and it was slow, but on Xen it was fast"?
Slashdot Mirror
To make this sort of test work, it must be run over a much longer period of time. But in the process of designing, building, testing and refining disk drive hardware and firmware (software), there isn't that much extra time to test drive failure rates. Want to wait an extra 9 months before releasing that new drive, to get accurate MTBF numbers? Didn't think so. How many different disk controllers do they use in the MTBF tests, to approximate different real-world behaviors? Probably not that many.
Could they run longer tests, and revise MTBF numbers after the initial release of a drive? Sure, and many of them do, but that revised MTBF would almost always be lower, making it harder to sell the drives. On the other hand, newer drives are certainly available every quarter, so it may not be a bad idea to lower the apparent value of older drive models.
So, it's better to assume a drive will fail before you're done using it. They're mechanical devices with high-speed moving parts, very narrow tolerable ranges of operation (that drive head has to be far enough away from the platters not to hit them, but close enough to read smaller and smaller areas of data). Anyone who's worked in a data center, or even a small server room, knows that drives fail. When I've had around two hundred drives, of varying ages, sizes and manufacturers, in a data center, I observed a failure rate of five to ten drives per year. This is well below the MTBF for enterprise disk array drives (SCSI, FC, SAS, whatever), but drives fail. That's why we have RAID. Storage Review has a good overview of how to interpret MTBF values from drive manufactures.
I think Chess is right - the Cone of Silence originated with Get Smart and was completely useless. I mean, it's a joke, and Mission: Impossible wasn't (intentionally) funny, though they did use the idea. But calling it a staple is an overstatement - laugh tracks, harried husbands, adorable children and perfect housewives were staples of 60's TV. Not like our modern, thought-provoking entertainment products...
Sure, it will still be a state-issued driver's license (or non-license ID). And many states already meet the pre-license verification and anti-counterfeiting measures required by RealID. But the bigger problem is that it all goes to a "secure," consolidated Federal database. What are the safeguards against abuse, tampering, or theft? DHS won't say. How are they maintaining this massive data warehouse? Nobody knows.
Try not to do anything suspicious, ok?
Careful on that slope buddy. I've been there, it's slippery. As long as all the people you interview are the same, in terms of discrimination, you're in good shape. But if some are older, or different genders, or different religions, or different ethnicities, you could set yourself and your company up for an employment discrimination lawsuit.
"But no," you say, "I'm just filtering for best fit." And that might work, more or less, but you pretty quickly come across people who are different from you, and if you turn them away because, let's say, they don't get your jokes, or they're not funny, and they happen to be a woman/Muslim/Asian/person-over-50... Well, what's the difference between discriminating based on personality (highly subjective) and discriminating based on race (also highly subjective)?
Now, age/sex/religious discrimination lawsuits don't happen that often in hiring situations, but if you're considering candidates for internal promotion, it gets pretty sticky.
Of course, fundamentally, the problem I see with your approach is that it boils down to hiring people you like. And in that pool of interviewees, people you've just met, the ones you like are the ones most like you. So you end up hiring people like you, turning away people who are different, and this hurts your organization. You need diversity of opinions, ideas, and personal experience to solve problems. Does your company want you to be comfortable, or productive? OK, team cohesion is important, but only because it helps performance - it's a means, not an end. Sure, you can get by with an army of clones, but you'll be doing yourself and your company a disservice.
But look at all innovative, closed-source operating system software running the world now!
VMWare ESX - based on Linux kernel with incremental enhancements
Windows NT/2000/XP/2003/2008 - designed by DEC VMS architects to use a microkernel architecture. Remind you of anything? Something that rhymes with "tunics"?
MS DOS - closed source software which appropriated most of its design from CP/M...
IBM OS/2 - DOS with a GUI and protected-mode memory access
IBM System z9 - UNIX, based on zSeries, based on System/390, based on System/370, based on System/360...
Mac OS X - UNIX, based on BSD
HP UX - UNIX
Wow! What a rich array of newness and innovation! Thanks, closed source software!
This is one of the worst networking articles I've ever read. There is nothing remotely interesting or informative in those two pages. There's not enough detail to offer any insight into networks, project management, system design or anything else. Why?
Why? Why? Why? Why? Why? Why? Why? Why? Why? Why? Why? Why? Why? Why? Why?
That's oversimplifying the service offerings and associated costs.
Verizon didn't lower their service charges by ~80%, undercutting their profitable T1 sales to business and institutional customers. They can only make more money by adding customers. FIOS primarily competes with cable-modem ISP services. It's a good deal compared to a cable modem and for a business which can't justify the cost of a T1, it looks like a good service.
But FIOS for business does not provide an SLA equivalent to Internet T1 service.
From the Verizon Business FIOS disclaimer:
Speed and uninterrupted use of the service are not guaranteed.
This is one of the main reasons why FIOS for business starts at $69/month and a Verizon (or similar) T1 with Internet service starts at around $600/month. The upload speed, average latency, and guaranteed latency are important factors as well.
In terms of variable costs, the SLA is probably the most expensive part of the service in terms of labor and management/monitoring infrastructure. Verizon has plenty of infrastructure to manage and monitor T1's, ISDN lines, POTS lines and other high-volume telecom circuit types. But FIOS uses different distribution, cable plant, and customer premises equipment than all the other Verizon services. This creates an entire new set of costs. This is why they don't offer it for FIOS - it's too expensive to build out the management systems, add NOC staff, and build the new FIOS network services at the same time.
FIOS may be offered with more substantial service guarantees in the future - I hope it is - but right now Verizon is learning how to operate their FIOS networks with consumers and small businesses as guinea pigs.
Sure, some are paid more than their productivity or merit might dictate, but the vast majority are underpaid, overworked, and without sufficient resources. I have yet to meet a teacher who doesn't purchase materials for their classroom (forget reimbursement). Nobody congratulates a new teacher on their personal-finance acumen. Communities rarely have enough money to maintain manageable class sizes (i.e. 20 or fewer students per teacher). But our political leaders and experts talk endlessly about how important children and education are - our future, etc. The teachers unions have a point - once you start differentiating between academic subjects, you create subjective differences all over the place. School system budgeting will become even more difficult to understand and time consuming than it is. You can create endless rules to encourage better teaching, but that won't improve the pool of applicants for teaching jobs.
So, double their pay. Across the board. Not just math and science teachers, but everyone.
"Oh no!" you say, "then we'll just be paying the bad teachers and the good teachers!" Well, true, but so what? Teachers are retiring at record rates. Perhaps you've heard of this "baby boomer" generation? Qualified, talented young teachers leave teaching because they don't get the pay/support/supplies they need. We need more teachers, and we need better teachers. Supply, meet demand.
Blizzard is already doing this. My company uses colo space in two facilities that also house Blizzard servers. I don't know the voltage, but all of their World of Warcraft server farms (~1000 blades each, around a dozen sites) run HP blades on DC power. The data centers have serious AC power distribution systems, but they've built custom DC power distribution in their floor space. The grounding wires - they're not small.
Cooling is provided by the facility's HVAC system, it's paid for (it's hard to bill each customer for hvac when they're all in a room the size of a football field). So Blizzard seems to be using DC to reduce total power consumption. The other benefit, as the parent suggested, is reduced wear on servers when they're not dealing with heat produced by inefficient AC/DC conversion. Of course, high-end power supplies are getting more and more efficient, but there's always some power loss/heat generation.
Even though exploits and bugs will continue to surface long after MS stops releasing patches, you'll save yourself a lot of risk if you log in as an unprivileged user. If you use the NTFS file system (and fixacls.exe if you have to convert after the install) your general-purpose login will have very few ways to wreck the system. Sure, there are privilege escalation attacks, but you'll be protected from many common bugs and exploits.
Keep a local admin account, or two, using a strong password. Change the default Administrator to a different user name and only use the admin priv's for maintenance.
Good luck.
Despite the study's flaws, many of the posts in this thread seem to focus on the inherent, perhaps necessary, superiority of Linux sysadmins to Windows sysadmins. This tangential argument actually makes the Microsoft argument. In fact, this has been the essential MSFT sales pitch for about 20 years - "our products are easier to use" - regardless of the competition. Sure, they sometimes try "our products are better", but the ease-of-use argument is their bread and butter.
The perceived, and fairly real, problem with Linux administration is that you have to do some level of programming. You don't have a choice. Most open-source (and commercial!) Linux applications require some level of configuration by hand, most commercial (and open source!) Windows applications do not. Since that manual configuration takes time, I suspect it diminishes the overall productivity of the skilled Linux admin (with requisite programming skills).
Can a Linux admin/developer really manage more servers than a Windows admin? My experience is that they can't - it tends to balance out. A well-configured, properly sized Windows server and a well-configured Linux server will both tend to run for quite a while without much intervention. Sure, many Linux applications are more memory and processor efficient than their Windows competition, and for large-scale environments this may be a decisive factor. But for most environments, even with dozens or hundreds of servers, the flexibility of a system in which you can modify anything, down to the kernel, doesn't provide enough business value to offset the high salary costs for sysadmins-with-programming-skills compared to those without. For people who don't know how to write shell scripts, or perl, or PHP - i.e. most of the world - Linux looks harder to use because it *is* harder to use.
Don't get me wrong - the open source approach is much more elegant - I'm just not sure it provides enough cost justification for most businesses. By arguing that Linux admins are required to be smarter and, incidentally, better paid than Windows admins, you're arguing against Linux adoption. Keep making this argument - Microsoft PR thanks you.
How often have you heard that the new version of Windows is "more secure" than the last version? A quick recap:
Windows 3.1 - no real security, but it's prettier than DOS!
Windows for Workgroups 3.1.1 - now with a login screen (but still no real security)!
Windows NT 3.51 - now with ACL's (and mostly not compatible with Win3.1 apps)!
Windows 95 - also has a login screen! no real security, but prettier than WfW!
Windows NT 4.0 - now with shared ACL's (domains) - the most secure Windows ever!
Windows 98 - Slightly less likely to crash than Win95! No NT security features!
Windows ME - Now with some system-software protection, but still no ACL's!
Windows 2000 - An improved interface and kernel! Active Directory 1.0! Now, the most secure Windows ever!
Windows XP - The successor to the Win2k and Win9x kernel products - super duper secure! Home users still run as the super-user, but it's less likely to crash! ACL's for Professional users and a very limited firewall make this, yes, the most secure Windows ever!
Windows 2003 (server) - The XP kernel in a server! Hardly anything runs by default! The Most Secure Windows Ever!
Windows Vista - Still with ACLs! New ways to limit access! Everyone's running as superuser, but with more warnings!
Windows Longhorn (server) - Not fully designed, but looks a little less secure than Win2003 - possibly *not* the most secure Windows ever!
So, this is an opinion piece, tying together loosely related facts, but there isn't much to worry about. Look at the product lines at these two companies - they make all kinds of non-AV software. These large ISV's get more insight into MSFT product road-maps than just about anybody - they knew this was coming.
I'm not sure where all the hatred for AV vendors comes from. Sure, you can't run a Windows system without an AV package, but that's not the fault of the AV vendors. These guys have been filling a need for DOS and Windows users for about 20 years. You don't have to buy their software, but their existence isn't hurting anyone. There are plenty of free and smaller-circulation commercial AV packages out there if you don't like these two. Use them instead. But to say "these companies must die" - really, why again? Because their software doesn't always work? Isn't that true of all software? Because once an AV package "messed up your computer" and you had to reinstall Windows? Maybe. More likely, your computer was already broken and an AV software installation was the straw that broke the camel's back.
AV vendors deal with a lot of confused end-users because their apps get deep into the OS and can get stomped by bad device drivers, OS changes, and other 3rd party software. AV software is often the canary in the coal-mine, alerting you to some underlying problem (your video driver has a memory leak, somebody's installer rolled-back an OS file to a previous version, etc).
If you don't like AV software, don't run Windows. Oh, wait, you don't? So what is your point exactly?
And $200 per user after that...
w tobuy/pricing.mspx
See http://www.microsoft.com/windowsserver2003/sbs/ho
And on the prevelance of MS Access - that prevelance is from home-grown database applications, not so much 3rd party vendors who can include MSDE with their products at almost no cost (with a Dev software license). If you're going to hack together your own Access database, it doesn't really matter what you use for a file server. With a little practice, any decent Access db programmer could build the same thing in MySQL (and get a free db server in the deal, if you want) or OpenOffice Base.
Let's say there are around two million regular Slashdot readers. I'd say 45% of Slashdot readers are puppies. Cute, cuddly, Linux-friendly puppies.
But that's not what 99% of wireless devices do - they need some wires configured before you can do anything. Wireless networking needs a wired access point, wireless mouse and keyboard need USB or PS2 wires to the box, and so on.
There are some nice opportunities for infrared and bluetooth devices to communicate with PC's, but this is an unplug-and-pray operation. Some things work together, some things don't.
As far as the assertion that Bluetooth rises up to allow all of our devices to sync with one another and the operating system , allow is the key word in that sentence. The only painless bluetooth operation I've seen is between two cell phones from the same manufacturer. They could share contacts very nicely. But connecting different bluetooth devices, or a bluetooth-enabled device and a Windows or Linux box, is not painless or predictable. Mac's seem to do this a little bit better than Windows/Linux (no big surprise there).
I have two "media pc's" in my living room - they're both Shuttle SFF pc's running Windows and Linux. No TV. No VCR. DVD's play just fine on either PC, with better picture quality than the TV I gave away two years ago. I don't use a remote, but I don't have a stack of tv/vcr/dvd/stereo remotes to contend with, so it's not a big loss. I'm not running MythTV or a Windows Media Center OS, but I play DVD's and video from iTunes and other sources on the computers. These devices have replaced a TV and VCR in my home. If I wanted TV reception, I could add TV tuner cards and get an antenna or a cable TV subscription, but I'm not really interested in TV shows
Lots of people use Mac Mini's for this purpose - they're just not called "Media PC's" - but they're small and quiet. The new iMacs come with a remote control and a quality LCD - I'd happily watch movies on those boxes.
The problem with the media PC, the convergence device, and so on, is that it isn't one pretty, easy-to-track product like an iPod or a Walkman. It's MythTV or TiVo (a little stretch, but not a big one) or a generic Linux/Windows/Mac computer which is, very slowly, replacing TV as media-consumption device. The fact that the PC architecture is rapidly evolving and quite flexible makes it attractive to people with only "hobbyist" level technical skills and circumvents the marketing efforts of PC and consumer-electronics manufacturers. The companies selling Home Entertainment PC's haven't found the right mix for a huge number of people to embrace, so people roll their own. Sure, most people over 40 would rather have super-simple consumer-electronics devices which "just work" - good for them! People in my age group, 30-40, may be a little more experimental, and I think people under 30 who are comfortable with computers and the Internet have even less use for TV.
So I think the Media PC is here, and getting bigger, but it will be hard to chronicle accurately for a few more years while the hardware/software/consumer-electronics vendors figure out what people really want and how to sell it to them.
This happens all the time, with all antivirus software. Signature-based or heuristic or both, they make mistakes. This sounds like a small mistake - it's much worse when you trigger off a widely used application or OS file.
Sure, this is one of the dangers of any malware detection software, but how else would you do it? You need to identify dangerous software using the smallest possible identifier or behavior. Sure, you could have a whole virus-library of malicious code in your detection system, but you would use a ton of memory and disk space to do it and the product would be unusable.
Same for heuristics - you want to exercise, in a pseudo-processing environment, every function in every piece of software on your machine to make sure it doesn't try anything nasty? You would lose most of your cpu cycles to heuristic analysis, waiting several minutes for a program to start executing *your* instructions after the malware analysis completes. Good luck with that project.
With 95% of the world's computers running Windows, everybody needs antivirus software. Unfortunate, and largely MSFT's fault, but there we are. NAV, SAV, McAfee, Grisoft, Trend, Sophos, etc. - they all have their pro's and con's.
Sure, it would be great to have an efficient anti-virus system that didn't depend on too-simple heuristics or too-hard-to-maintain signatures. Does someone know how to do that? There's a big AV market out there. Go ahead, build it.
This is the whole problem with integrating consumer entertainment media (CD's, DVD's, etc.) and general-purpose computing devices. General-purpose hardware with general-purpose operating systems are good at lots and lots of things. That's why people buy billions of $US worth of computers and software every year. But these general-purpose devices don't fit the expected distribution conditions for entertainment media companies - they can't handle an environment in which the consumer can change the use of the media. The DVD and CD players work just fine in this model - they just play the media.
But the machine with the keyboard is trouble because of the keyboard and the general-purpose OS environment. Microsoft started down the road to DRM-friendly media-player-only operating systems with Windows Media Edition, but this version of Windows still maintains all the existing Windows functionality.
This could work out well if an enterprising OS vendor merged cheap PC's with a media-player operating system to create single-purpose appliances. But the general-purpose OS genie is out of the bottle and hardware isn't cheap enough yet (no $50 computers, plenty of $50 DVD players) for people to give up web-browsing and text editing and email so they can have a dedicated media-player. So these two worlds are joined and general-purpose computing is already suffering.
-don
As long as Intel maintains this architecture, with a single data bus for RAM, PCI, PCI-e, AGP, BIOS, and other integrated functions, they'll be behind AMD. AMD's current (and future) HyperTransport provides a wider, more efficient data path than the front-side bus. AMD's per-processor memory controller scales past two sockets in a way that Intel just can't match. By pushing fully-buffered DIMM's, each with its own memory controller, Intel is ceding the design point to AMD: a single memory controller is too much of a bottleneck, the load needs to be spread around. This is especially true when you go beyond two processors in a machine, but even dual-socket boxes benefit from distributed memory controllers. Sure, the Bensley FSB goes to 1033Mhz from 800Mhz, but that doesn't sound like a big jump.
Until Intel has a real answer to HyperTransport, they'll be losing the high-performance, 4+ sockets market to AMD. For smaller two-socket servers, Intel will have to pay the RAM and/or server vendors to make FB-DIMM's price competitive with different flavors of DDR.
I do some work in financial services, including working with SWIFT.
SWIFT is a network, not a database. It can store and forward data, but only for short periods of time.
SWIFT transactions can be done directly between organizations over the network, without stopping at SWIFT systems, making them slightly harder to track.
SWIFT is run by its 7,000+ member organizations, from all over the world. Most SWIFT members are banks, and the largest banks have the most influence among the members. So it's not US-based, but it's well stocked by US banks and corporations. It's not like the United Nations - the biggest players run the operation.
Since SWIFT is not a database, and not all transactions land on SWIFT computers, you can't really "query SWIFT" about financial data. You can only "query SWIFT" about the status of a message from party A or to party B. So to do data mining on SWIFT data, you would need to build your own SWIFT data warehouse and your own (massive) matrix of data relationships.
SWIFT is a network, like the Internet, but entirely private. The problem of gathering data from SWIFT communications is similar to the challenge in gathering data from AT&T's Internet backbone routers. But filtering the Internet data is easier - you can probably drop bit-torrent traffic, ISO's, binary software packages, iTunes media, etc. That's a lot of packets to ignore. If you're building a comprehensive financial database from uncorrelated data, in a multitude of formats and languages, you have to start by recording everything just to make some sense of the data.
Also, why would terrorist groups use banks? For the convenience? Certainly, wire transfers and ATM's are the *easiest* way to move money around the world. But if you were trying not to leave tracks, you probably would avoid banks or use one-time-use bank accounts for individual money transfers. People have been solving this problem in organized crime for decades, why wouldn't the same methods be available to Bin Laden or anyone else?
Strong words, but pretty accurate.
I manage a few small data centers and we depend on RAID 1 and RAID 5 (and redundant servers) to keep our business running. Down-time is expensive, but rebuilding a machine from scratch is expensive too. So we don't use any software RAID or software-assisted on-board "controllers" - it's hard to call them controllers. RAID Filters? RAID bridges? RAID-like adapters? I love real RAID controllers. Everybody I interview has to explain RAID and something about how it works. If someone tells me they use RAID-0 for performance, they'll need a really impressive story about why that's not plain stupid.
People like to say "RAID-0 is great for gamers and testers - it's cheap and fast," but think about the scenarios where this applies:
1. Cheap!!
You have one computer, you use it (mostly) for gaming. Your critical data exists elsewhere (GMail? cd-rw's?). A second, identical hard drive runs you about $100 US. You have on-board RAID, which relies on your processor for all the heavy lifting, but it's paid for already. When you're loading your multi-gigabyte game, the two spindles go pretty fast together and your processor, while not idle, is mostly waiting for the disks. So you can spare the CPU cycles the RAID "controller" uses, great. But this computer is twice as likely to crash and burn as a single-HDD machine, so you can't rely on it to hold any persistent data (personal finance, schoolwork, download-only software, etc, etc). You just diminished the value of your computer, and you're trying to save money.
2. Fast, but just for games!!
You have a dedicated computer, with two-disk RAID-0, and it's your gaming rig. It's optimized for speed. The on-board RAID controller isn't fast enough when you've got to load textures from the hard drive and track hundreds of independant 3d objects in (insert blow-up-stuff game title here). So you have a hardware-based controller, with its own BIOS and OS-independant management tools. You could buy three RAID cards like this for the cost of a "mainstream" video card, the cost isn't a big deal - you can afford to spend $2k on an unstable drag-racer of a gaming rig.
Someone in the world thinks your time isn't worthless, but you're willing to burn a whole day, now and then, to rebuild your computer? Why not spend another ~$200, get two more hard drives (I know you've got the fans/power supply/drive bays/liquid-nitrogen-cooling to handle it), and run RAID 1+0? You get the same speed boost, but now you don't need ***a whole other computer*** to store your "real" data.
Don't get me wrong, big, resource-intensive games are great stuff. And faster loading is great too. But how many computers do you really want to own and maintain? I know, for many people, having six machines to manage them is cooler than one, but my time is more valuable than that (I know, we're mostly talking about Windows, and you can maintain thousands of *nix boxes with perl-scripts and ssh, thanks, got it). I'd rather just get a faster drive and controller - SATA2 is pretty damn fast for a non-SCSI system.
3. Fast for testing!! No real data!!
You're not testing performance, right? Because you're not depending on RAID-0 to handle real data when the test is complete... If you have a grid computing farm, and the time to rebuild a single box is only a few minutes, and you can stand some data loss, this might be a good solution. Maybe your application can withstand individual machine failures and the raw speed is more important... In this kind of setup, you're much more likely to have network bottlenecks than disk bottlenecks -- your processor-intensive tasks are passed around the network and not too much data has to be stored on disks. But if this is the case, and you can afford a grid/server farm, I doubt you're going to solve a disk-IO bottleneck with RAID 0 instead of external storage (SAN, big SCSI array, big SATA/SAS array, etc).
There probably are situations where the additional data-loss risk is offset by performance gains, but usual pro-RAID-0 scenarios in this debate just don't make sense.
PS - mod parent up!
The Mozilla suite, now SeaMonkey, has had user-profile sync'ing to FTP servers for about a year. SeaMonkey 1.0.2 has roaming profile functions to HTTP or FTP servers as a native feature.
-Don
What sort of testing did you do? Was it something like "3 linux DNS servers with 10k concurrent requests per minute for 60 minutes on different platforms"? Or more like "I loaded Ubuntu on ESX and it was slow, but on Xen it was fast"?
-Don
Who run Bartertown now? And when will they build the Thunderdome?
-DC