This is a good thing, the B53 was a last ditch weapon intended to take out the hardened bunkers of the Soviet leadership, except it was air burst which is a highly, highly ineffective was to take out a bunker. The replacement is a much smaller, much less dirty penetrator weapon, the B61 Mod 11.
Dude, where have you been TB is still here and is scarier than ever because most cases are now resilient to all but a cocktail of the most infrequently used antibiotics with the most severe side effects.
Considering they're Jag's the owners are already very familiar with the garage so no need for something like WiFi =)
Hell most of them will be updated during the next visit, not because they received the notification.
If you're worried about paying off your loan you probably won't qualify for 5-6 figures worth of cash advances... Hell I make about 2x the median houshold income, have near perfect credit, and I only have five to ten thousand worth of cash advance limit on my 3 cards combined.
Actually your library records are no longer covered, that part of the PATRIOT Act was not renewed. FISA only covers foreign agents. I know because the head librarian for the Medina County Library system has been one of the champions against COPA, the Patriot Act provision, and any other legislation which restricts the freedom of her patrons to freely read what they want without the government peering over their shoulder.
That better be all read cache, otherwise the first time you lose power or have a kernel panic you're losing a LOT of data. That's what I was talking about with the need to disable write cache.
Not all of those are atom based, and even the ones that are probably are using embedded optimized designs not general purpose desktop boards, and they have correctly sized power supplies. Can you build a similar project by sourcing the same components? Sure you can, but for cost and availability reasons most home NAS systems that I have seen are just desktop boards thrown in a midsized tower case with a desktop PSU.
The other nice thing about ZFS is L2ARC and ZIL, throw in one or two cheap SSD's and your TB's of cheap storage start to perform like a 5-6 figure array =)
And the small and midsized businesses that are the main audience for these devices don't care, they don't want to be good at fixing obscure problems with LVM, they want a device to hold their data which requires the least amount of staff time to setup and care for.
No, RAID6 is the only useful level of RAID for any 7200 RPM drive over ~1TB other than RAID10. The bit error rate and time to rebuild are too high for anyone who cares about their data to use anything else.
And the homebrew NAS also costs more in time to setup, has no support, and uses probably anywhere from a few dozen to a few hundred watts more power. Basically the majority of the people interested in the tested solutions would not consider a Linux box with some storage to be a viable alternative to those boxes.
Btw I'm very surprised at the performance of the StorCenter px6, considering before the device launched they used it at EMC World to boot 100 VDI machines in like a minute and a half with SSD's I have to assume the performance limitation was the bundled drives. I wonder why EMC/Iomega crippled the retail bundle with such low performing drives knowing it would be benchmarked against the competition like this?
I'm not saying more lanes isn't nice, or useful. I was specifically responding to but just shy of the magical 48 you'd need for 3x fully populated pci express x16 slots. Giving each of the physical x16 slots an electrical x8 connection is not going to hamper any existing peripheral that I am aware of. With the doubling of per lane bandwidth from PCIe 1 to PCIe 2 the need for an electrical x16 slot was basically eliminated. (if it was ever needed in the first place, was there ever a device that needed more than 2GB/s that was fitted with a PCIe 1.x chipset?)
Nothing needs 3x PCIe2.0 slots let alone PCIe3! No graphics card is hampered by having 8x lanes, nor is there any external I/O card capable of saturating 8x lanes. Heck even the $15k PCIe flash cards I've used in my datacenter aren't hampered by 8x PCIe 2.0.
Like I said, officially it's not there. I have little doubt that it's just a single command away from being turned on, but I doubt it would do any good. The reason is that with GLONASS nearing completion, Galileo coming online, and differential GPS you aren't going to be able to meaningfully confuse the positioning system on anything you're really worried about.
FT is so limiting as to be useless in all but a handful of situations. I was hoping many of the restrictions would be lifted with vsphere 5 but that did not happen so it's pretty obvious to me that SMP synchronization with the method they are using is a tough nut to crack.
Seagate and Hitachi make the only actually reliable HDD's, but they're a LOT more expensive than consumer drives. However even the expensive enterprise class drives have an AFR of ~1.5% so the only solution is RAID (and with modern capacity relative to IOPS RAID10 and RAID6 are the only real options due to extended rebuild times).
Sprint has it out of the box on their Android phones but it requires a $30/month wireless tethering plan! This is in addition to the $10/month extra for it being a smartphone!!
You'll get PSOD's, heck just recently there was one caused by upgrading your vcenter server to version 5 while you still had 4.0 U2 hosts. Given the various problems with 4.1 there are quite a few people who might be upgrading directly (we will be) and so run into that bug (we upgraded to 4.0 U3 to avoid it, but only because I saw a blog post and so was forewarned). Also my main point was that you don't want a flaky whitebox or otherwise unreliable host so VMWare or any other hypervisor is not a magic bullet that allows you to go for unreliable, ultra cheap hardware (quite the opposite IMHO as you have more eggs in the one basket).
Exactly, when you lose a host through a hardware or software fault all the VM's that were running on it have to be restarted on other hosts in the cluster. Also to anyone with a brain the fact that I mention vmotion and VMWare HA would mean I'm obviously not running a Windows host =)
Slashdot Mirror
This is a good thing, the B53 was a last ditch weapon intended to take out the hardened bunkers of the Soviet leadership, except it was air burst which is a highly, highly ineffective was to take out a bunker. The replacement is a much smaller, much less dirty penetrator weapon, the B61 Mod 11.
Was $510 the cost of just the shot(s) or did that also include the doctors office visits?
Dude, where have you been TB is still here and is scarier than ever because most cases are now resilient to all but a cocktail of the most infrequently used antibiotics with the most severe side effects.
Of course running a nuclear plant is cheap if you can externalize risk....
Considering they're Jag's the owners are already very familiar with the garage so no need for something like WiFi =)
Hell most of them will be updated during the next visit, not because they received the notification.
If you're worried about paying off your loan you probably won't qualify for 5-6 figures worth of cash advances...
Hell I make about 2x the median houshold income, have near perfect credit, and I only have five to ten thousand worth of cash advance limit on my 3 cards combined.
Actually your library records are no longer covered, that part of the PATRIOT Act was not renewed. FISA only covers foreign agents. I know because the head librarian for the Medina County Library system has been one of the champions against COPA, the Patriot Act provision, and any other legislation which restricts the freedom of her patrons to freely read what they want without the government peering over their shoulder.
That better be all read cache, otherwise the first time you lose power or have a kernel panic you're losing a LOT of data. That's what I was talking about with the need to disable write cache.
Not all of those are atom based, and even the ones that are probably are using embedded optimized designs not general purpose desktop boards, and they have correctly sized power supplies. Can you build a similar project by sourcing the same components? Sure you can, but for cost and availability reasons most home NAS systems that I have seen are just desktop boards thrown in a midsized tower case with a desktop PSU.
The other nice thing about ZFS is L2ARC and ZIL, throw in one or two cheap SSD's and your TB's of cheap storage start to perform like a 5-6 figure array =)
And the small and midsized businesses that are the main audience for these devices don't care, they don't want to be good at fixing obscure problems with LVM, they want a device to hold their data which requires the least amount of staff time to setup and care for.
No, RAID6 is the only useful level of RAID for any 7200 RPM drive over ~1TB other than RAID10. The bit error rate and time to rebuild are too high for anyone who cares about their data to use anything else.
And the homebrew NAS also costs more in time to setup, has no support, and uses probably anywhere from a few dozen to a few hundred watts more power. Basically the majority of the people interested in the tested solutions would not consider a Linux box with some storage to be a viable alternative to those boxes.
Btw I'm very surprised at the performance of the StorCenter px6, considering before the device launched they used it at EMC World to boot 100 VDI machines in like a minute and a half with SSD's I have to assume the performance limitation was the bundled drives. I wonder why EMC/Iomega crippled the retail bundle with such low performing drives knowing it would be benchmarked against the competition like this?
I'm not saying more lanes isn't nice, or useful. I was specifically responding to but just shy of the magical 48 you'd need for 3x fully populated pci express x16 slots. Giving each of the physical x16 slots an electrical x8 connection is not going to hamper any existing peripheral that I am aware of. With the doubling of per lane bandwidth from PCIe 1 to PCIe 2 the need for an electrical x16 slot was basically eliminated. (if it was ever needed in the first place, was there ever a device that needed more than 2GB/s that was fitted with a PCIe 1.x chipset?)
Nothing needs 3x PCIe2.0 slots let alone PCIe3! No graphics card is hampered by having 8x lanes, nor is there any external I/O card capable of saturating 8x lanes. Heck even the $15k PCIe flash cards I've used in my datacenter aren't hampered by 8x PCIe 2.0.
Like I said, officially it's not there. I have little doubt that it's just a single command away from being turned on, but I doubt it would do any good. The reason is that with GLONASS nearing completion, Galileo coming online, and differential GPS you aren't going to be able to meaningfully confuse the positioning system on anything you're really worried about.
FT is so limiting as to be useless in all but a handful of situations. I was hoping many of the restrictions would be lifted with vsphere 5 but that did not happen so it's pretty obvious to me that SMP synchronization with the method they are using is a tough nut to crack.
SA is officially dead via presidential order. It's not even available in the most recent version of bird (at least officially).
Only if you can live without write cache
Which controllers let you do RAID1 across 3 drives? I believe with the HP controllers you can only do paired drives in RAID1 or groups of 2 in RAID10.
Exactly, or anywhere in the US midwest either. 2 feet of snow might shut down the city for a day depending on when it comes relative to rush hour.
Seagate and Hitachi make the only actually reliable HDD's, but they're a LOT more expensive than consumer drives. However even the expensive enterprise class drives have an AFR of ~1.5% so the only solution is RAID (and with modern capacity relative to IOPS RAID10 and RAID6 are the only real options due to extended rebuild times).
Sprint has it out of the box on their Android phones but it requires a $30/month wireless tethering plan! This is in addition to the $10/month extra for it being a smartphone!!
You'll get PSOD's, heck just recently there was one caused by upgrading your vcenter server to version 5 while you still had 4.0 U2 hosts. Given the various problems with 4.1 there are quite a few people who might be upgrading directly (we will be) and so run into that bug (we upgraded to 4.0 U3 to avoid it, but only because I saw a blog post and so was forewarned). Also my main point was that you don't want a flaky whitebox or otherwise unreliable host so VMWare or any other hypervisor is not a magic bullet that allows you to go for unreliable, ultra cheap hardware (quite the opposite IMHO as you have more eggs in the one basket).
Exactly, when you lose a host through a hardware or software fault all the VM's that were running on it have to be restarted on other hosts in the cluster. Also to anyone with a brain the fact that I mention vmotion and VMWare HA would mean I'm obviously not running a Windows host =)