SMP will only bring you so far - i'll bet 8 VCPU VMs on Atoms will be beat by a 2 VCPU VM on a Core 2 Duo.
Perhaps not, depending on the other load the system is working on. Because of the way VCPUs are scheduled (at least in VMWare) that 8-vCPU VM won't get a time-slice until such time as there are 8 real cores available for the duration of that slice. If your task is CPU intensive and can be easily separated into distinct tasks not overly chatty (i.e. cross VM latency is not going to be a major issue) and the host has gobs of RAM available, you are often better off having several VMs with one cVPU each than one VM with several vCPUs. This may be much less of a problem on a many-CPU monster like the 512 core unit being discussed than it is on 2/4/8-core boxes, but I expect the balance to still be in favour of multiple single-vCPU VMs in cases where the task can be efficiently split between them.
It appears many of you have a horrible misunderstanding of how the VMware hypervisor works. One instance of VMware ESX can be installed on a single system image machine. SSI is defined by shared memory. There are two classes of systems with shared memory: SMP and NUMA. This 512 Atom machine is neither. It is a system of 512 individual machines cobbled together with very neat thermally and electrically efficient packaging. If one were to run VMware ESX on this machine, it would require 512 instances of the VMware ESX kernel. You could then assign guests to each single Atom machine, but you would not be able to schedule guests across/among the 512 machines. VMware, or any virtualization would be useless on this platform. That's not what it's designed for. To get efficient compute bound applications to run effectively on this machine would require something like MOSIX. For other applications such as web or webmail, you'd use a standard load balancer and a a cluster of 512 individual servers, just as is done now.
Again, this machine is not SMP and not NUMA. It is not a "server" in the traditional sense. In reality it is a _cluster_ in a box. The concept is nothing new, this company is rather later to the game, and is already beat in the efficiency department. One example beating it is the Sicortex SC5832 which houses 5832 64 bit MIPS cores running at 700MHz. The CPU die contains 6 MIPS64 cores, dual memory controllers, a PCIe interface, GigE interface, and a connection to the Kautz digraph network running at 2.5GB/s bidirectional to from each node. The system consumes less than 20KW for almost 6000 cores and over 8Tflops of floating point performance. This highly integrated cluster machine is designed for supercomputing, not web applications, though it could be used for such if desired.
The reason you can't come up with any examples of natural monopolies is that there aren't any.
You need to read up on your history.
Microsoft. AT&T. Standard Oil. US Steel.
All natural monopolies, each more "natural" than the previous one. All busted by the government for their anti-competitive practices.
The prosecution of U.S. Steel failed, so they weren't proven a monopoly. AT&T was a monopoly created by US law, thus not a 'natural' monopoly. AT&T didn't engage in anti competitive practices as Congress set the rates and profit structure until breaking AT&T up in the early 1980s. Your other 3 examples all engaged in such practices. Standard Oil and Microsoft were both prosecuted under the Sherman antitrust act. Standard Oil was found guilty and broken up. Microsoft settled to avoid breakup.
Meaning all patents and designs of those products are now owned by the public. The public overpaid for them so force the companies to give something back to the public. The manufacturing processes and techniques can be kept secret but all the chip design and patents should be open for competitors to step in and make a better cheaper product.
You didn't think this through before posting. DRAM is a commodity. It's traded on the markets just like orange juice and wheat. No electronics producers care who makes the DRAM that goes into their products. They buy on price alone, because all the chips are the same--i.e. they're a commodity. No company controls a patent on DRAM today. That's why it's so damn cheap, and that's why these companies engaged in price fixing--BECAUSE none of them has a patent allowing them to control the price INDIVIDUALLY. The only current patent holder that can control any price aspect of DRAM is RAMBUS, because they pulled a sneaky and got JEDEC to adopt their signaling interface for DDR2/3. And, RAMBUS doesn't produce products, only patents. Rambus wasn't involved in the price fixing.
The reason DRAM is cheap is the LACK of patents. Sprinkle some greed on top of this low profit cut throat market segment, and there you have the price fixing. Thus, your "surrender the patents" suggestion won't gain anyone anything as no "3rd party" will rush in and produce DRAM more cheaply. There is no efficiency in production left to be gained. There's no more profit to be had. DRAM profit was cut to the bone ten years ago, and as a result these producers have been engaging in various levels of price fixing pretty much for the entire decade.
regardless of why, I've never heard of that happen with an ext3 filesystem. Now imagine you're running a server, a trip to the datacentre to run fsck would be annoying.
Both created to solve this exact problem. Both work very well. IPMI is included in the standard config of many servers, and is thus "free". iKVM solutions can get very pricey.
Btfrs already seems to be more stable than ext4: every PC I own with an ext4 partition has failed to boot at some point due to disk corruption
This is what a ~100MB/boot partition, typically formatted with EXT2, is for. A boot environment doesn't need a journaled FS due to its small size and the fact that it's typically only written to when a new kernel is installed. Journal? WTF for?
Oh, and use LILO, because Grub, in all it's glorious versions, sucks. Grub attempts to be a Swiss army knife of bootloaders for all archs, when all most people ever need is the spoon. I've been peeved since the day Debian switched to Grub as the default bootloader. Makes for problems for LILO users. I'll use LILO til it's pulled from my cold dead hands.
I've run filesystems that were considered ok-but-early-adopter on servers before. Early XFS releases, for example. It's perhaps not really comparable as SGI had already developed XFS v1 on their workstations and so most of the code was fairly heavily-tested before the Linux port of XFS v2..
XFS was, and is, very mature. However, I just want to point out that XFS gained most of its scalability, maturity, and performance development whilst running on the Origin and Onyx 2000/3000 supercomputers than it ever did on SGI's workstation platforms such as the Octane and O2 systems. SGIs supercomputer platforms became the bulk of their system sales shortly after introduction, with workstation sales falling off dramatically from the late 1990s into the early 2000s then eventually being dropped completely, along with all the other vendors' RISC workstations, including DEC, IBM, HP, and SUN.
With the Origin/Onyx architecture and its XIO bus chip fully integrated into the NUMALink interconnect came Guaranteed Rate I/O code in IRIX and the Real Time option in XFS. The XIO hub chip contained hardware logic that allowed guaranteed data rates for a given filesystem at the bus hardware level. It was absolutely ingenious technology, and allowed SGI to dominate the digital broadcast video business for quite some time. Unfortunately GRIO didn't make it into the Altix systems or Linux as XIO was dropped in favor of PCI-x and PCIe in order to lower costs and increase the availability of third party HBAs.
Regarding XFS scalability, SGI's sole focus from the late 90s on was the scientific supercomputer/server market. From the Origin/Onyx 2000 onward, SGI's NUMA/IRIX machines scaled initially to 512 sockets in a single system image, with NASA Ames successfully running an O2K system with 1024 sockets in SSI. These machines needed massive amounts of RAID storage for high speed scratch space and semi-permanent data storage, and XFS had to provide high performance filesystem operations atop the IRIX volume manager spanning dozens if not hundreds of external FC SAN RAID LUNs. XFS was providing Terabyte filesystems before most of us had ever heard the word "terabyte". Today, on SGI hardware at various labs around the world, XFS is supporting petabyte filesystems.
Yeah right. Celebrities won't be forced to go through these things.
Mouthy arrogant celebs will--the ones who proactively say "I'm _not_ going through that thing!". That's an automatic response of: "In you go, or to jail you go."
Also, any male TSA grunt running the scanner will force the likes of Scarlett Johansson, Megan Fox, et al to go through it. That's just guaranteed.
Sounds like classic bipolar mania or psychosis. The state of mental health care in the US is pretty abysmal for those without health insurance or any ability to pay for decent care. What's the state of mental health diagnosis and treatment in Russia? I'd bet it's abysmal, period. This cat has probably done and said many wacky things throughout his life/career, but until now, none of them were outlandish enough to gain any kind of real notice. Or, maybe he just never showed any symptoms until the Vodka ran out. Alcohol is a natural self medicating device for many with mental illness--helps mask the symptoms. There is also anecdotal evidence showing well above average IQ in the bipolar population, though no formal study has shown quantification AFAIK.
Sun's service has been sliding for some time now. Oracle appears to be accelerating that decline. We had some RAIDs, originally purchased from StorageTek before the Sun acquisition, come off of the three year warranty they were purchased with. We've been unable to get Sun (now Oracle) to recognize the RAID's serial numbers to get them on the maintenance contract for quite some time now. You'd think Oracle would want our money?
Not for kit that old. They'd prolly rather sell you a new array, which their people can actually support. If they even have spare parts for it they probably can't locate them. How many years ago did Sun acquire StorageTek? 5? How old is that pre-acquisition array of yours? At least 5+ years old. Given its age and limited capacity/performance, may I ask why you're still using it? And given the plethora of quality cheap storage arrays on the market, such as Nexsan http://www.nexsan.com/ why would you not just replace that StorageTek array with such a unit instead of continuing your masochist ways with "lock in" vendors such as Sun and Oracle?
Yes, Palin only said that you could see Russia from Alaska, which is only a hair less idiotic when you consider that she was trying to claim that as a reason for why she has experience in international politics. Most people don't differentiate because both comments are so idiotic that there isn't a difference worth caring about.
Well, this proves the first statement is not idiotic, as you put it, but is factual.
"At their closest Alaska and Russia are 2.5 miles apart – the distance between Little Diomede Island, Alaska, and Big Diomede Island, Russia. The two islands straddle the U.S.-Russian maritime border in the middle of the Bering Strait. In mid-winter, when the Bering Strait freezes, it is possible to walk between the two islands – from American to Russia, from today to tomorrow, or from Russia to the United States, from today to yesterday. It is even possible to stand on the frozen Bering Strait, with one foot in America and one foot in Russia, straddling the frontiers of distant boundaries and time travel."
"In the north, the date line swings to the east through the Bering Strait, and then west past the Aleutian Islands to keep Alaska (part of the United States) and Russia, which are due north and south of each other in that region, on opposite sides of the line and in agreement with the date in the rest of those countries. As a result of this line-adjusting, the right to call itself "The Last Place on Earth" (that is, the latest place) goes to the westernmost Aleutian Island of Attu.
The date line passes equidistantly between the two Diomede Islands—Little Diomede Island (US) and Big Diomede Island (Russia)—at a distance of 1.5 km (1 mi) from each island."
Palin's foreign policy experience remark regarding the proximity of Russia was a humorous rebuttal and counterattack on Obama's lack of foreign policy experience. Neither of them had any. Palin was saying, humorously, that she has slightly more foreign policy experience than Obama because Russia is right next door to Alaska. This remark was twisted by the liberal mainstream media, as usual, to make a Republican candidate look bad. If you have seen the video of that rally, it is abundantly clear she was poking fun at the opposition, _not_ making a serious statement about her foreign affairs experience.
Nah, we need a good old fashioned global thermonuclear war. It would kill billions of evil humans whom we've been told are the cause of global warming, and it would create a nuclear winter, causing worldwide temperatures to plummet by 50C throwing us into an ice age for a thousand years. This would eliminate the global warming problem and we evil instigators of it in one push of the button.
No one at the company had the forethought to do something like this. The same could also be accomplished merely by purchasing 20 identical Dells or HPs. Every serious vendor offers business platforms for a 2-3 year period, to solve or prevent this exact problem. The problem was purchasing dissimilar hardware to begin with. Now you're paying the price. *Next time* do everything you can to push $employer to buy identical machines. Explain the reasons for doing so. Use the data at the Intel and AMD sites above to make your case.
It seems Oracle is explicitly disinterested in Java, so IBM may get the one thing they would have wanted on the cheap, a chance at the people behind Sun's Java as they leave/are forced out of Oracle.
IBM has a chance if it lets these folks stay in the bay area and telecommute, or opens an office for them there. Environment plays a huge factor in people's lives and outlook, and moving from San Fran to say Yorktown Heights, NY probably wouldn't give these folks warm fuzzies. They'd probably prefer to stay in the bay area. Thus, I agree with a previous OP that this guy and others will likely end up at Google, or other bay area firms. It's called Silicon Valley for a reason. There are many companies there who could benefit from the talent that will be fleeing the sinking SUN ship. Both Intel and Apple are in the valley, as well as HP and others. Speaking of HP, now that Oracle has SUN's big iron, HP is the only one of the remaining general market big 3 Unix hardware vendors without its own DB engine. Oracle now has SPARC/OracleDB, IBM has i/p/zSeries and DB2, HP has SuperDome (Itanium) and offers OracleDB on top of HP UX or MS SQL Server on top Windows Datacenter Edition for Itanium.
Now that Oracle will have its own big iron to push for OracleDB, look for HP to acquire and build its own enterprise database engine or lavish resources upon PostgreSQL and push it. I'm sure up to this point HP has been just fine reselling Oracle and MSSQL. The acquisition of SUN is going to change HPs push of OracleDB atop HP UX, whether HP wants that or not. Ellison is going to push Oracle atop SPARC.
Then again, printers and ink cartridges make up something like 60% of HP's revenue, and servers and storage something like only 10%, so they may not really care if they can sell Unix iron in the future with a top notch DB atop.
Database software is now finally catching up with postgreSQL. You can argue that mysql is getting better too I suppose. With the database software formally reserved to big iron we have no reason to use Solaris/Oracle except on all but the biggest data warehouses.
That is not correct. Linux far out scales Solaris due to the limitations of the largest Sun server. Linux efficiently scales to a higher processor and memory count than any other *nix OS on the planet. Why? SGI. You can now purchase a packaged SGI Altix UV 1000 system with 256 sockets, each containing an 8 core Xeon 7500 CPU for a total of 2048 cores and 16TB of shared memory in 4 standard racks tied together by the NUMALink 5 interconnect with 40GB/s bidirectional interconnect bandwidth to each two socket blade, all running a single copy of standard SuSE Linux.
The largest Sun system, the M9000, tops out at 64 sockets each with a 4 core Fujitsu SPARC64 VII CPU for a total of 256 cores and 4TB of shared memory. On a per core basis the Xeon 7500 is well over 50% faster than the SPARC64 VII and the Altix has 8 times as many cores.
Either of these systems has enough PCIe slots to connect more adapters, bandwidth and storage than any database could ever need.
Assuming PostgreSQL can efficiently fork and/or thread and can create and manage multi terabyte database files, this SGI Altix UV system running SuSE Linux and PostgreSQL would run circles around any other single system available. All with an essentially free Linux distro and FOSS DB. Don't like SuSE? You could install Debian or any other Linux on the Altix UV, although SGI probably wouldn't support it. The NUMA magic is already in the mainline Linux kernel, so any Linux distro, properly configured, could run on this Altix system.
So, you can get a FOSS Linux OS and FOSS DB and likely beat anything on the planet. The cost of the hardware would make you choke, however. It's up in the $2+ million range without storage. It would still be cheaper than the SUN M9000 + OracleDB because the OS and DB are in essence free. You'd automatically pay for a SuSE support contract as SGI won't sell the box(es) without it. You'd want it anyway though, as things do go wrong, and especially on a box with 2048 cores. You'd probably want a support contract for PostgreSQL as well. But, both these contracts will be less than Solaris+Oracle.
Even the "midrange" Altix UV 100 would kick the crap out of SUN's biggest box. The Altix UV 100 is a two rack system with 96 Xeon 7500 CPUs, 768 cores, and 6TB of memory. Cheaper than its big brother, the UV 1000, but probably still over $1 million without storage.
It's too bad SGI only focuses on the scientific supercomputing market. These new Altix UV systems, especially since based on the x86-64 Xeon instead of Itanium, would just absolutely trounce many traditional Unix systems in the large database market, as well as many other markets.
That ended several years ago. Fujitsu/Amdahl exited the mainframe market not long after z/Architecture became available. They never resold IBM hardware; they built their own, and decided that the market couldn't justify the expense they'd incur trying to keep up plus the patent royalties IBM was demanding.
My mistake. I didn't know it was a royalty situation. So Fujitsu was rolling their own compatible hardware under a patent license.
There are zero platform choices available to someone who can't spend hundreds of millions to get rid of their mainframes. They are locked in, due entirely to IBM's actions. If a company makes decisions based on there being competition in the market, and then one company eliminates the competition, that's monopolistic tactics.
One "platform" doesn't make a "market". The IBM mainframe platform comprises less than 1% of installed server platforms by type. By unit volume it's a tiny fraction of 1%. One cannot be deemed a monopoly when one owns less than 1% of a "market". Platform != Market. Monopoly only applies to markets, not individual platforms. Again, the mere fact that it's extremely expensive to migrate doesn't by itself denote a monopoly. What would create a monopoly is the total lack of the choice to migrate.
I don't care for IBM's mainframe business practices, but they're nowhere close to the definition of "monopolistic" under the Sherman Antitrust Act.
They do have mainframe competition to an extent. But the competition are not true competitors, they're IBM partners, i.e. OEM customers. Or, at least there used to be. I'm thinking Fujitsu/Ahmdal. They compete somewhat with IBM for the sale, but the electronics complex in those machines is True Blue IIRC, supplied in an OEM fashion by IBM, under an Iron clad contract of course. IBM makes a little less on the hardware and software, but this relationship has kept world-wide antitrust regulators at bay for some decades. IIRC Siemens sells (or sold) a rebadged Fujitsu mainframe, twice removed from IBM, but still containing all IBM electronics inside.
Thus, IBM does have some mainframe competition, FSVO "competition". Or at least they used to as of a few years ago. I tend to look at this like the courts do. There is plenty of technological choice in the marketplace today WRT to platforms. If people choose to stick with IBM mainframe lock in, that's their choice. That doesn't make IBM a monopolist, because there are hundreds of other platform choices available. Allowing oneself to become locked in in the first place is not a result of "monopoly" but bad decision making. If a company truly want to get out from under IBM's proverbial mainframe lock in thumb, all they have to do is make the choice to spend the money and port/convert/migrate. The fact that it's expensive to migrate is also not a result of a "monopoly".
Mainframes are still the guy hidden in the shadows, smoking the cigarette; he's still there and has more power than you think.
The only "power" you speak of is vendor lock in due to BS contract language. The only thing really holding IBM's mainframe division together, for the most part, for the past 20 years, are their lawyers and the lock in contracts they write for the mainframe platforms.
A shot across the bow is a bad thing, given their current position and that of people that respect FOSS. You'd think they'd have given this more thought.
Indeed. Big Blue has made far more $$ off Linux consolidation on zSeries than they can possibly lose due to the few customers who move off zSeries hardware to this emulation platform. Those who would do so will do it because they simply can't afford the IBM lock in pricing any longer. Those customers aren't typically buying or leasing new mainframes anyway. They limp along forever, putting off the inevitable switch to another platform, because IBM is always willing to offer an extended full covered service contract at a "just reasonable enough" price, no matter how old the mainframe hardware in question may be.
If this Hercules emulator is such a threat, why doesn't Big Blue pull the standard maneuver of "buying" the project, people, and Hercules the company, bring it all under the roof, and fully support it? THAT is smart business, not wasting money tying to kill the competition. This Hercules solution would pick off the weak and injured mainframe customers at most. It is not a threat to IBM's healthy mainframe customers as they usually need the advanced mainframe features not available in the emulator on x86 hardware solution.
I feel they're on the path to losing more money due to FOSS backlash than what they'd save by suing this "competition" into non existence.
Ultra-Lightweight cars were attempted before. You crash, you die.
No, not at all. Indy cars, for example, are vastly lighter than any standard American cars, and they crash at extremely high speeds with very few fatalities, and often without even injuries to the driver.
When's the last time you saw an average motorist strapped into his/her car with a 5 point harness and a racing helmet, which is now locked to the seat back on each side by a strap, prompted across all racing bodies by the whiplash death of Dale Earnhart?
Put a standard shoulder belt and airbag in an Indy car and see what happens in these crashes. Death, every time. You're comparing apples to oranges. The entire safety package of an Indy car makes it relatively safe for drivers crashing at high speeds. Only a portion of that safety is provided by the collapsing chassis.
Slashdot Mirror
The ads IN THE PAPER would be targeted using geolocation from the IP address of your printer so that you would get locally appropriate ads.
What's the geolocation of 192.168.x.x?
SMP will only bring you so far - i'll bet 8 VCPU VMs on Atoms will be beat by a 2 VCPU VM on a Core 2 Duo.
Perhaps not, depending on the other load the system is working on. Because of the way VCPUs are scheduled (at least in VMWare) that 8-vCPU VM won't get a time-slice until such time as there are 8 real cores available for the duration of that slice. If your task is CPU intensive and can be easily separated into distinct tasks not overly chatty (i.e. cross VM latency is not going to be a major issue) and the host has gobs of RAM available, you are often better off having several VMs with one cVPU each than one VM with several vCPUs. This may be much less of a problem on a many-CPU monster like the 512 core unit being discussed than it is on 2/4/8-core boxes, but I expect the balance to still be in favour of multiple single-vCPU VMs in cases where the task can be efficiently split between them.
It appears many of you have a horrible misunderstanding of how the VMware hypervisor works. One instance of VMware ESX can be installed on a single system image machine. SSI is defined by shared memory. There are two classes of systems with shared memory: SMP and NUMA. This 512 Atom machine is neither. It is a system of 512 individual machines cobbled together with very neat thermally and electrically efficient packaging. If one were to run VMware ESX on this machine, it would require 512 instances of the VMware ESX kernel. You could then assign guests to each single Atom machine, but you would not be able to schedule guests across/among the 512 machines. VMware, or any virtualization would be useless on this platform. That's not what it's designed for. To get efficient compute bound applications to run effectively on this machine would require something like MOSIX. For other applications such as web or webmail, you'd use a standard load balancer and a a cluster of 512 individual servers, just as is done now.
Again, this machine is not SMP and not NUMA. It is not a "server" in the traditional sense. In reality it is a _cluster_ in a box. The concept is nothing new, this company is rather later to the game, and is already beat in the efficiency department. One example beating it is the Sicortex SC5832 which houses 5832 64 bit MIPS cores running at 700MHz. The CPU die contains 6 MIPS64 cores, dual memory controllers, a PCIe interface, GigE interface, and a connection to the Kautz digraph network running at 2.5GB/s bidirectional to from each node. The system consumes less than 20KW for almost 6000 cores and over 8Tflops of floating point performance. This highly integrated cluster machine is designed for supercomputing, not web applications, though it could be used for such if desired.
http://www.sicortex.com/
http://www.sicortex.com/registration/download/265/1886/file/TechSummary_FINAL.pdf
The reason you can't come up with any examples of natural monopolies is that there aren't any.
You need to read up on your history.
Microsoft. AT&T. Standard Oil. US Steel.
All natural monopolies, each more "natural" than the previous one. All busted by the government for their anti-competitive practices.
The prosecution of U.S. Steel failed, so they weren't proven a monopoly. AT&T was a monopoly created by US law, thus not a 'natural' monopoly. AT&T didn't engage in anti competitive practices as Congress set the rates and profit structure until breaking AT&T up in the early 1980s. Your other 3 examples all engaged in such practices. Standard Oil and Microsoft were both prosecuted under the Sherman antitrust act. Standard Oil was found guilty and broken up. Microsoft settled to avoid breakup.
I believe some of those culprit corps are Japanese, Mitsubishi & Toshiba at least.
Germany -----Infineon
Japan ----------NEC, Hitachi, Mitsubishi, Toshiba, Elpida
S. Koea -------Samsung, Hynix
Taiwain --------Nanya
USA ------------Micron
Meaning all patents and designs of those products are now owned by the public. The public overpaid for them so force the companies to give something back to the public. The manufacturing processes and techniques can be kept secret but all the chip design and patents should be open for competitors to step in and make a better cheaper product.
You didn't think this through before posting. DRAM is a commodity. It's traded on the markets just like orange juice and wheat. No electronics producers care who makes the DRAM that goes into their products. They buy on price alone, because all the chips are the same--i.e. they're a commodity. No company controls a patent on DRAM today. That's why it's so damn cheap, and that's why these companies engaged in price fixing--BECAUSE none of them has a patent allowing them to control the price INDIVIDUALLY. The only current patent holder that can control any price aspect of DRAM is RAMBUS, because they pulled a sneaky and got JEDEC to adopt their signaling interface for DDR2/3. And, RAMBUS doesn't produce products, only patents. Rambus wasn't involved in the price fixing.
The reason DRAM is cheap is the LACK of patents. Sprinkle some greed on top of this low profit cut throat market segment, and there you have the price fixing. Thus, your "surrender the patents" suggestion won't gain anyone anything as no "3rd party" will rush in and produce DRAM more cheaply. There is no efficiency in production left to be gained. There's no more profit to be had. DRAM profit was cut to the bone ten years ago, and as a result these producers have been engaging in various levels of price fixing pretty much for the entire decade.
I swear. One day there will be a FS-related post on slashdot WITHOUT a Reiser-joke.
Not today.
How many Russian brides does it take to kill a filesystem?
regardless of why, I've never heard of that happen with an ext3 filesystem. Now imagine you're running a server, a trip to the datacentre to run fsck would be annoying.
Apparently you've never heard of IPMI: http://en.wikipedia.org/wiki/Intelligent_Platform_Management_Interface
or
iKVM: http://en.wikipedia.org/wiki/KVM_switch
Both created to solve this exact problem. Both work very well. IPMI is included in the standard config of many servers, and is thus "free". iKVM solutions can get very pricey.
Btfrs already seems to be more stable than ext4: every PC I own with an ext4 partition has failed to boot at some point due to disk corruption
This is what a ~100MB /boot partition, typically formatted with EXT2, is for. A boot environment doesn't need a journaled FS due to its small size and the fact that it's typically only written to when a new kernel is installed. Journal? WTF for?
Oh, and use LILO, because Grub, in all it's glorious versions, sucks. Grub attempts to be a Swiss army knife of bootloaders for all archs, when all most people ever need is the spoon. I've been peeved since the day Debian switched to Grub as the default bootloader. Makes for problems for LILO users. I'll use LILO til it's pulled from my cold dead hands.
I've run filesystems that were considered ok-but-early-adopter on servers before. Early XFS releases, for example. It's perhaps not really comparable as SGI had already developed XFS v1 on their workstations and so most of the code was fairly heavily-tested before the Linux port of XFS v2..
XFS was, and is, very mature. However, I just want to point out that XFS gained most of its scalability, maturity, and performance development whilst running on the Origin and Onyx 2000/3000 supercomputers than it ever did on SGI's workstation platforms such as the Octane and O2 systems. SGIs supercomputer platforms became the bulk of their system sales shortly after introduction, with workstation sales falling off dramatically from the late 1990s into the early 2000s then eventually being dropped completely, along with all the other vendors' RISC workstations, including DEC, IBM, HP, and SUN.
With the Origin/Onyx architecture and its XIO bus chip fully integrated into the NUMALink interconnect came Guaranteed Rate I/O code in IRIX and the Real Time option in XFS. The XIO hub chip contained hardware logic that allowed guaranteed data rates for a given filesystem at the bus hardware level. It was absolutely ingenious technology, and allowed SGI to dominate the digital broadcast video business for quite some time. Unfortunately GRIO didn't make it into the Altix systems or Linux as XIO was dropped in favor of PCI-x and PCIe in order to lower costs and increase the availability of third party HBAs.
Regarding XFS scalability, SGI's sole focus from the late 90s on was the scientific supercomputer/server market. From the Origin/Onyx 2000 onward, SGI's NUMA/IRIX machines scaled initially to 512 sockets in a single system image, with NASA Ames successfully running an O2K system with 1024 sockets in SSI. These machines needed massive amounts of RAID storage for high speed scratch space and semi-permanent data storage, and XFS had to provide high performance filesystem operations atop the IRIX volume manager spanning dozens if not hundreds of external FC SAN RAID LUNs. XFS was providing Terabyte filesystems before most of us had ever heard the word "terabyte". Today, on SGI hardware at various labs around the world, XFS is supporting petabyte filesystems.
.
To get truly astonishing pictures, they should add a black pixel, to improve contrast.
Light switches and window shades are cheaper, and probably more effective.
Yeah right. Celebrities won't be forced to go through these things.
Mouthy arrogant celebs will--the ones who proactively say "I'm _not_ going through that thing!". That's an automatic response of: "In you go, or to jail you go."
Also, any male TSA grunt running the scanner will force the likes of Scarlett Johansson, Megan Fox, et al to go through it. That's just guaranteed.
Why do high profile chess players always have to go completely batshit crazy?
Sounds like classic bipolar mania or psychosis. The state of mental health care in the US is pretty abysmal for those without health insurance or any ability to pay for decent care. What's the state of mental health diagnosis and treatment in Russia? I'd bet it's abysmal, period. This cat has probably done and said many wacky things throughout his life/career, but until now, none of them were outlandish enough to gain any kind of real notice. Or, maybe he just never showed any symptoms until the Vodka ran out. Alcohol is a natural self medicating device for many with mental illness--helps mask the symptoms. There is also anecdotal evidence showing well above average IQ in the bipolar population, though no formal study has shown quantification AFAIK.
...Turbo switches on our workstations again like back in the day?
Whadya mean 'back in the day'?
Sun's service has been sliding for some time now. Oracle appears to be accelerating that decline. We had some RAIDs, originally purchased from StorageTek before the Sun acquisition, come off of the three year warranty they were purchased with. We've been unable to get Sun (now Oracle) to recognize the RAID's serial numbers to get them on the maintenance contract for quite some time now. You'd think Oracle would want our money?
Not for kit that old. They'd prolly rather sell you a new array, which their people can actually support. If they even have spare parts for it they probably can't locate them. How many years ago did Sun acquire StorageTek? 5? How old is that pre-acquisition array of yours? At least 5+ years old. Given its age and limited capacity/performance, may I ask why you're still using it? And given the plethora of quality cheap storage arrays on the market, such as Nexsan http://www.nexsan.com/ why would you not just replace that StorageTek array with such a unit instead of continuing your masochist ways with "lock in" vendors such as Sun and Oracle?
Yes, Palin only said that you could see Russia from Alaska, which is only a hair less idiotic when you consider that she was trying to claim that as a reason for why she has experience in international politics. Most people don't differentiate because both comments are so idiotic that there isn't a difference worth caring about.
Well, this proves the first statement is not idiotic, as you put it, but is factual.
From: http://www.gov.state.ak.us/trade/2003/tad/russia/facts.htm
"At their closest Alaska and Russia are 2.5 miles apart – the distance between Little Diomede Island, Alaska, and Big Diomede Island, Russia. The two islands straddle the U.S.-Russian maritime border in the middle of the Bering Strait. In mid-winter, when the Bering Strait freezes, it is possible to walk between the two islands – from American to Russia, from today to tomorrow, or from Russia to the United States, from today to yesterday. It is even possible to stand on the frozen Bering Strait, with one foot in America and one foot in Russia, straddling the frontiers of distant boundaries and time travel."
If you don't get the part about "today to tomorrow" and vice versa, see: http://en.wikipedia.org/wiki/International_Date_Line
"In the north, the date line swings to the east through the Bering Strait, and then west past the Aleutian Islands to keep Alaska (part of the United States) and Russia, which are due north and south of each other in that region, on opposite sides of the line and in agreement with the date in the rest of those countries. As a result of this line-adjusting, the right to call itself "The Last Place on Earth" (that is, the latest place) goes to the westernmost Aleutian Island of Attu.
The date line passes equidistantly between the two Diomede Islands—Little Diomede Island (US) and Big Diomede Island (Russia)—at a distance of 1.5 km (1 mi) from each island."
Palin's foreign policy experience remark regarding the proximity of Russia was a humorous rebuttal and counterattack on Obama's lack of foreign policy experience. Neither of them had any. Palin was saying, humorously, that she has slightly more foreign policy experience than Obama because Russia is right next door to Alaska. This remark was twisted by the liberal mainstream media, as usual, to make a Republican candidate look bad. If you have seen the video of that rally, it is abundantly clear she was poking fun at the opposition, _not_ making a serious statement about her foreign affairs experience.
You mean something like volcano ash, but higher?
Hmm... maybe we need a bigger volcano...
Nah, we need a good old fashioned global thermonuclear war. It would kill billions of evil humans whom we've been told are the cause of global warming, and it would create a nuclear winter, causing worldwide temperatures to plummet by 50C throwing us into an ice age for a thousand years. This would eliminate the global warming problem and we evil instigators of it in one push of the button.
This is actually what he *needs*:
http://www.intel.com/cd/business/enterprise/emea/eng/189154.htm
http://www.amd.com/gb-uk/Processors/ProductInformation/0,,30_118_14397,00.html
No one at the company had the forethought to do something like this. The same could also be accomplished merely by purchasing 20 identical Dells or HPs. Every serious vendor offers business platforms for a 2-3 year period, to solve or prevent this exact problem. The problem was purchasing dissimilar hardware to begin with. Now you're paying the price. *Next time* do everything you can to push $employer to buy identical machines. Explain the reasons for doing so. Use the data at the Intel and AMD sites above to make your case.
It seems Oracle is explicitly disinterested in Java, so IBM may get the one thing they would have wanted on the cheap, a chance at the people behind Sun's Java as they leave/are forced out of Oracle.
IBM has a chance if it lets these folks stay in the bay area and telecommute, or opens an office for them there. Environment plays a huge factor in people's lives and outlook, and moving from San Fran to say Yorktown Heights, NY probably wouldn't give these folks warm fuzzies. They'd probably prefer to stay in the bay area. Thus, I agree with a previous OP that this guy and others will likely end up at Google, or other bay area firms. It's called Silicon Valley for a reason. There are many companies there who could benefit from the talent that will be fleeing the sinking SUN ship. Both Intel and Apple are in the valley, as well as HP and others. Speaking of HP, now that Oracle has SUN's big iron, HP is the only one of the remaining general market big 3 Unix hardware vendors without its own DB engine. Oracle now has SPARC/OracleDB, IBM has i/p/zSeries and DB2, HP has SuperDome (Itanium) and offers OracleDB on top of HP UX or MS SQL Server on top Windows Datacenter Edition for Itanium.
Now that Oracle will have its own big iron to push for OracleDB, look for HP to acquire and build its own enterprise database engine or lavish resources upon PostgreSQL and push it. I'm sure up to this point HP has been just fine reselling Oracle and MSSQL. The acquisition of SUN is going to change HPs push of OracleDB atop HP UX, whether HP wants that or not. Ellison is going to push Oracle atop SPARC.
Then again, printers and ink cartridges make up something like 60% of HP's revenue, and servers and storage something like only 10%, so they may not really care if they can sell Unix iron in the future with a top notch DB atop.
Database software is now finally catching up with postgreSQL. You can argue that mysql is getting better too I suppose. With the database software formally reserved to big iron we have no reason to use Solaris/Oracle except on all but the biggest data warehouses.
That is not correct. Linux far out scales Solaris due to the limitations of the largest Sun server. Linux efficiently scales to a higher processor and memory count than any other *nix OS on the planet. Why? SGI. You can now purchase a packaged SGI Altix UV 1000 system with 256 sockets, each containing an 8 core Xeon 7500 CPU for a total of 2048 cores and 16TB of shared memory in 4 standard racks tied together by the NUMALink 5 interconnect with 40GB/s bidirectional interconnect bandwidth to each two socket blade, all running a single copy of standard SuSE Linux.
The largest Sun system, the M9000, tops out at 64 sockets each with a 4 core Fujitsu SPARC64 VII CPU for a total of 256 cores and 4TB of shared memory. On a per core basis the Xeon 7500 is well over 50% faster than the SPARC64 VII and the Altix has 8 times as many cores.
Either of these systems has enough PCIe slots to connect more adapters, bandwidth and storage than any database could ever need.
Assuming PostgreSQL can efficiently fork and/or thread and can create and manage multi terabyte database files, this SGI Altix UV system running SuSE Linux and PostgreSQL would run circles around any other single system available. All with an essentially free Linux distro and FOSS DB. Don't like SuSE? You could install Debian or any other Linux on the Altix UV, although SGI probably wouldn't support it. The NUMA magic is already in the mainline Linux kernel, so any Linux distro, properly configured, could run on this Altix system.
So, you can get a FOSS Linux OS and FOSS DB and likely beat anything on the planet. The cost of the hardware would make you choke, however. It's up in the $2+ million range without storage. It would still be cheaper than the SUN M9000 + OracleDB because the OS and DB are in essence free. You'd automatically pay for a SuSE support contract as SGI won't sell the box(es) without it. You'd want it anyway though, as things do go wrong, and especially on a box with 2048 cores. You'd probably want a support contract for PostgreSQL as well. But, both these contracts will be less than Solaris+Oracle.
Even the "midrange" Altix UV 100 would kick the crap out of SUN's biggest box. The Altix UV 100 is a two rack system with 96 Xeon 7500 CPUs, 768 cores, and 6TB of memory. Cheaper than its big brother, the UV 1000, but probably still over $1 million without storage.
It's too bad SGI only focuses on the scientific supercomputing market. These new Altix UV systems, especially since based on the x86-64 Xeon instead of Itanium, would just absolutely trounce many traditional Unix systems in the large database market, as well as many other markets.
http://www.sgi.com/products/servers/altix/uv/
That ended several years ago. Fujitsu/Amdahl exited the mainframe market not long after z/Architecture became available. They never resold IBM hardware; they built their own, and decided that the market couldn't justify the expense they'd incur trying to keep up plus the patent royalties IBM was demanding.
My mistake. I didn't know it was a royalty situation. So Fujitsu was rolling their own compatible hardware under a patent license.
There are zero platform choices available to someone who can't spend hundreds of millions to get rid of their mainframes. They are locked in, due entirely to IBM's actions. If a company makes decisions based on there being competition in the market, and then one company eliminates the competition, that's monopolistic tactics.
One "platform" doesn't make a "market". The IBM mainframe platform comprises less than 1% of installed server platforms by type. By unit volume it's a tiny fraction of 1%. One cannot be deemed a monopoly when one owns less than 1% of a "market". Platform != Market. Monopoly only applies to markets, not individual platforms. Again, the mere fact that it's extremely expensive to migrate doesn't by itself denote a monopoly. What would create a monopoly is the total lack of the choice to migrate.
I don't care for IBM's mainframe business practices, but they're nowhere close to the definition of "monopolistic" under the Sherman Antitrust Act.
They do have mainframe competition to an extent. But the competition are not true competitors, they're IBM partners, i.e. OEM customers. Or, at least there used to be. I'm thinking Fujitsu/Ahmdal. They compete somewhat with IBM for the sale, but the electronics complex in those machines is True Blue IIRC, supplied in an OEM fashion by IBM, under an Iron clad contract of course. IBM makes a little less on the hardware and software, but this relationship has kept world-wide antitrust regulators at bay for some decades. IIRC Siemens sells (or sold) a rebadged Fujitsu mainframe, twice removed from IBM, but still containing all IBM electronics inside.
Thus, IBM does have some mainframe competition, FSVO "competition". Or at least they used to as of a few years ago. I tend to look at this like the courts do. There is plenty of technological choice in the marketplace today WRT to platforms. If people choose to stick with IBM mainframe lock in, that's their choice. That doesn't make IBM a monopolist, because there are hundreds of other platform choices available. Allowing oneself to become locked in in the first place is not a result of "monopoly" but bad decision making. If a company truly want to get out from under IBM's proverbial mainframe lock in thumb, all they have to do is make the choice to spend the money and port/convert/migrate. The fact that it's expensive to migrate is also not a result of a "monopoly".
I didn't say buy and kill. I said "buy" (FSVO "buy") and then fully support the solution.
Mainframes are still the guy hidden in the shadows, smoking the cigarette; he's still there and has more power than you think.
The only "power" you speak of is vendor lock in due to BS contract language. The only thing really holding IBM's mainframe division together, for the most part, for the past 20 years, are their lawyers and the lock in contracts they write for the mainframe platforms.
A shot across the bow is a bad thing, given their current position and that of people that respect FOSS. You'd think they'd have given this more thought.
Indeed. Big Blue has made far more $$ off Linux consolidation on zSeries than they can possibly lose due to the few customers who move off zSeries hardware to this emulation platform. Those who would do so will do it because they simply can't afford the IBM lock in pricing any longer. Those customers aren't typically buying or leasing new mainframes anyway. They limp along forever, putting off the inevitable switch to another platform, because IBM is always willing to offer an extended full covered service contract at a "just reasonable enough" price, no matter how old the mainframe hardware in question may be.
If this Hercules emulator is such a threat, why doesn't Big Blue pull the standard maneuver of "buying" the project, people, and Hercules the company, bring it all under the roof, and fully support it? THAT is smart business, not wasting money tying to kill the competition. This Hercules solution would pick off the weak and injured mainframe customers at most. It is not a threat to IBM's healthy mainframe customers as they usually need the advanced mainframe features not available in the emulator on x86 hardware solution.
I feel they're on the path to losing more money due to FOSS backlash than what they'd save by suing this "competition" into non existence.
Ultra-Lightweight cars were attempted before.
You crash, you die.
No, not at all. Indy cars, for example, are vastly lighter than any standard American cars, and they crash at extremely high speeds with very few fatalities, and often without even injuries to the driver.
When's the last time you saw an average motorist strapped into his/her car with a 5 point harness and a racing helmet, which is now locked to the seat back on each side by a strap, prompted across all racing bodies by the whiplash death of Dale Earnhart?
Put a standard shoulder belt and airbag in an Indy car and see what happens in these crashes. Death, every time. You're comparing apples to oranges. The entire safety package of an Indy car makes it relatively safe for drivers crashing at high speeds. Only a portion of that safety is provided by the collapsing chassis.