The unique failure mode for SSDs is related to the NAND substrate wearing out. However, modern SSDs detect this and rewrite their data before it's at risk of being uncorrectable. While a brand new drive may hold data for 100 years if you've only written it once, a drive at the end of warranty having been used to the max may only hold data in your closet for a year before it starts showing errors on readback. Just watch your SMART tables and you should be fine, since SSDs tell you how many write cycles are remaining as a percentage.
The opposite of that is the failure mode of most hard drives which is a head crash either due to mechanical damage or ESD. Once that happens, your entire drive is dead, unless you have many thousands of dollars for a recovery service.
The above assumes that the chance of failure due to things like PCBA solder failures etc. is roughly common and similar across both types of products.
Not to be rude, but I'm guessing there's only 20-30 engineers in the world who have any idea what current wear leveling state-of-the-art is, and how it affects performance.
There's a huge variety in the quality of controllers in the marketplace, and just because one design has a given advantage or flaw, doesn't mean others share those attributes.
You should *not* say "he did some stupid things in the past" because that will open you up to a lawsuit if said person can ever track that comment back to you. It's way too vague, and probably none of their business. Screening candidates accurately isn't your job, it's theirs.
The safest things to say to an HR cold call regarding an applicant are either glowing recommendations or "Sorry, but I have no feedback to offer on the person you're asking about."
Residential power here is $0.055 per kWh with no hourly, daily or seasonal restrictions or pricing... roughly 25% of what you pay. There are almost no cases where paying extra for a green technology locally will recover the costs.
It's been tried, and it's basically too expensive as a single device... today's hybrid drives can't afford enough flash (or fit it on their boards) to be very fast, and it complicates manufacturing immensely.
The solution is motherboards with more SATA ports (original ICH6 had two, now we see 6-8 SATA ports on normal consumer motherboards) so you can have one of each, an SSD and a rotating drive, each manufactured in the most efficient way possible.
Further, SSDs have evolved enough that you really don't need to worry about wearing out the devices in all but the most extreme workloads.
ZFS will even manage your desired hierarchy automatically by attaching an SSD as a cache device to a zpool, keeping frequently read content in the cache device and flushing periodically any changes through to the rotating media.
I think this kills the market for people who want to goof off at work, as there's no way those SC2 packets will make it through a corporate firewall, but they'd probably do just fine going back and forth with your cubicle neighbor.
I think we read different articles. He's not saying he didn't plan well enough, he's saying that Intel and AMD promise that Gen Y processor is 35% faster than Gen X processor, and he's not seeing anywhere near 35% in real world performance.
If the application was purely CPU bound, and Y wasn't giving me 35% more than X, I'd complain.
However, if it's a complex system like almost everything else, why would they expect their application to get 35% faster when there's probably 6 or 8 critical subsystems that could all be bottlenecks as well?
Intel, AMD, Sun, HP, Dell, IBM, whoever. They all provide a wide range of configurable, scalable, inexpensive (to expensive) building blocks to get your business need accomplished. With all the amazing things that have been created using computers in the last thirty or forty years, I can't blame the providers of those building blocks for something not coming together. The latest chips are absolutely faster than previous ones, but maybe you need to rethink your network fabric? Maybe your storage subsystem? Maybe your application design? Who knows where their problem really lies.
If anyone could throw together a few white boxes for next to nothing and build something to survive facebook's traffic volumes, then there wouldn't be any jobs for smart IT folks, would there? Google seems to have figured something out, Amazon seems to have figured something out, maybe Facebook needs to apply more brains to their problem instead of complaining.
He's probably too young to get a "real" job according to many state labor laws.
When I was a kid, you had to be 14.5 to get a work permit, and could only work 20 hours/week. At 15 you could work 28 hours/week, and at 16 you could then work full-time (40 hours).
The issue isn't overall voter support in Pennsylvania, as he would get plenty of votes from people who registered as Democrats.
The problem is that he wouldn't survive a republican primary in his own state, since the voting in that primary would more heavily favor a right-of-center candidate than the overall voter biases would indicate.
If he doesn't win the party primary, it's tough to be elected, since I don't believe Pennsylvania allows people to run as independents if they lose the primary. (Unlike Connecticut and Lieberman)
Actually, the the X25-M is for "mainstream" usage, including laptops and desktops. The X25-E is for extreme workloads, including some server usages.
The X25-M is available both in 1.8" and 2.5" SATA form factors, which are the two most common laptop interfaces today.
PCIe is a bit more limited in a laptop typically, and if you go that route (as a laptop manufacturer) you're generally locking yourself into a single device vendor, since you'll need custom drivers for whichever PCIe board you choose. SATA, on the other hand, allows you to pick any device on the market if you follow the form factor guidelines properly.
Actually, SATA protocol is reasonably low overhead, compared to SAS/SCSI, I wouldn't say it's a massive limiter.
As long as you plan on allocating 1 PCIe bitlane per SSD attached, whether SATA or anything else, you should be fine. The problem is those PCIe x1 devices with four drives attached. They can't keep any of the SATA busses saturated.
However, get a good x4 controller to attach 4 devices, or x8 for 8 devices, and you'll be able to keep everything humming along at their peak rates.
Remember, a single modern rotating drive (130-140MB/s sequential at the OD) can saturate any regular PCI controller, and almost saturate an x1 PCIe controller by itself. It gets even worse when your NIC and your HBA share a PCI bus, because you can lose half your bandwidth going in-and-out of memory.
Problem with getting six IoDrives is having six PCI-e slots available that are all x4 or better. However, you can get that same IO rate in IOPS with a pair of high-end 16-port PCI-e SATA/SAS boards, and connect 32 Intel X25-E drives, and likely still have expansion slots left over in your system. That whole setup will burn about 5W per controller, and ~2W per drive under load.
Slashdot Mirror
actually it's physics...
The unique failure mode for SSDs is related to the NAND substrate wearing out. However, modern SSDs detect this and rewrite their data before it's at risk of being uncorrectable. While a brand new drive may hold data for 100 years if you've only written it once, a drive at the end of warranty having been used to the max may only hold data in your closet for a year before it starts showing errors on readback. Just watch your SMART tables and you should be fine, since SSDs tell you how many write cycles are remaining as a percentage.
The opposite of that is the failure mode of most hard drives which is a head crash either due to mechanical damage or ESD. Once that happens, your entire drive is dead, unless you have many thousands of dollars for a recovery service.
The above assumes that the chance of failure due to things like PCBA solder failures etc. is roughly common and similar across both types of products.
Or drinking copious amounts of booze while it's too frackin cold to go outside anyway.
Agreed. From TFA:
"had a growing queue of important engineering projects outside of our core technology"
outside the core technology is the important part
Not to be rude, but I'm guessing there's only 20-30 engineers in the world who have any idea what current wear leveling state-of-the-art is, and how it affects performance.
There's a huge variety in the quality of controllers in the marketplace, and just because one design has a given advantage or flaw, doesn't mean others share those attributes.
Following Forge's ideas are a bad idea IMO.
You should *not* say "he did some stupid things in the past" because that will open you up to a lawsuit if said person can ever track that comment back to you. It's way too vague, and probably none of their business. Screening candidates accurately isn't your job, it's theirs.
The safest things to say to an HR cold call regarding an applicant are either glowing recommendations or "Sorry, but I have no feedback to offer on the person you're asking about."
That, and I'm not sure how you produce hard evidence that something as vague as "competition problems" do not exist.
That's not interesting. How do you plan to connect a non-Intel CPU to an Intel chipset with integrated graphics?
Depends heavily on where you live.
Residential power here is $0.055 per kWh with no hourly, daily or seasonal restrictions or pricing... roughly 25% of what you pay. There are almost no cases where paying extra for a green technology locally will recover the costs.
http://www.youtube.com/watch?v=_v_azJv50KQ
Amazing what 4 hours of makeup and 4 hours of photoshop can do.
someone mod the above up please, the AC is exactly right.
It's been tried, and it's basically too expensive as a single device... today's hybrid drives can't afford enough flash (or fit it on their boards) to be very fast, and it complicates manufacturing immensely.
The solution is motherboards with more SATA ports (original ICH6 had two, now we see 6-8 SATA ports on normal consumer motherboards) so you can have one of each, an SSD and a rotating drive, each manufactured in the most efficient way possible.
Further, SSDs have evolved enough that you really don't need to worry about wearing out the devices in all but the most extreme workloads.
ZFS will even manage your desired hierarchy automatically by attaching an SSD as a cache device to a zpool, keeping frequently read content in the cache device and flushing periodically any changes through to the rotating media.
I think this kills the market for people who want to goof off at work, as there's no way those SC2 packets will make it through a corporate firewall, but they'd probably do just fine going back and forth with your cubicle neighbor.
I think we read different articles. He's not saying he didn't plan well enough, he's saying that Intel and AMD promise that Gen Y processor is 35% faster than Gen X processor, and he's not seeing anywhere near 35% in real world performance.
If the application was purely CPU bound, and Y wasn't giving me 35% more than X, I'd complain.
However, if it's a complex system like almost everything else, why would they expect their application to get 35% faster when there's probably 6 or 8 critical subsystems that could all be bottlenecks as well?
Mod the above up.
Intel, AMD, Sun, HP, Dell, IBM, whoever. They all provide a wide range of configurable, scalable, inexpensive (to expensive) building blocks to get your business need accomplished. With all the amazing things that have been created using computers in the last thirty or forty years, I can't blame the providers of those building blocks for something not coming together. The latest chips are absolutely faster than previous ones, but maybe you need to rethink your network fabric? Maybe your storage subsystem? Maybe your application design? Who knows where their problem really lies.
If anyone could throw together a few white boxes for next to nothing and build something to survive facebook's traffic volumes, then there wouldn't be any jobs for smart IT folks, would there? Google seems to have figured something out, Amazon seems to have figured something out, maybe Facebook needs to apply more brains to their problem instead of complaining.
Can't decide if this should be +5 insightful or +5 funny
Do you honestly feel ~2 years of grief and frustration is worth less than $200? Then by all means, don't cancel, stand up for your principles.
Any copy-on-write system needs a scratchpad area to be effective.
Were you unable to scale the disks a bit past your workload size?
He's probably too young to get a "real" job according to many state labor laws.
When I was a kid, you had to be 14.5 to get a work permit, and could only work 20 hours/week. At 15 you could work 28 hours/week, and at 16 you could then work full-time (40 hours).
how much do you pay for two tablespoons of beach sand?
Doubtful.
The issue isn't overall voter support in Pennsylvania, as he would get plenty of votes from people who registered as Democrats.
The problem is that he wouldn't survive a republican primary in his own state, since the voting in that primary would more heavily favor a right-of-center candidate than the overall voter biases would indicate.
If he doesn't win the party primary, it's tough to be elected, since I don't believe Pennsylvania allows people to run as independents if they lose the primary. (Unlike Connecticut and Lieberman)
The X25-M datasheet guarantees 20GB/day for 5 years. How many DVDs to you torrent each day?
I don't think it necessarily eliminates the value proposition...
The X25-E claims a petabyte of lifetime random writes, plus it's quite a bit faster.
There are applications for each out there, though you're right, the majority of users will be perfectly happy with the X25-M
Actually, the the X25-M is for "mainstream" usage, including laptops and desktops. The X25-E is for extreme workloads, including some server usages.
The X25-M is available both in 1.8" and 2.5" SATA form factors, which are the two most common laptop interfaces today.
PCIe is a bit more limited in a laptop typically, and if you go that route (as a laptop manufacturer) you're generally locking yourself into a single device vendor, since you'll need custom drivers for whichever PCIe board you choose. SATA, on the other hand, allows you to pick any device on the market if you follow the form factor guidelines properly.
Actually, SATA protocol is reasonably low overhead, compared to SAS/SCSI, I wouldn't say it's a massive limiter.
As long as you plan on allocating 1 PCIe bitlane per SSD attached, whether SATA or anything else, you should be fine. The problem is those PCIe x1 devices with four drives attached. They can't keep any of the SATA busses saturated.
However, get a good x4 controller to attach 4 devices, or x8 for 8 devices, and you'll be able to keep everything humming along at their peak rates.
Remember, a single modern rotating drive (130-140MB/s sequential at the OD) can saturate any regular PCI controller, and almost saturate an x1 PCIe controller by itself. It gets even worse when your NIC and your HBA share a PCI bus, because you can lose half your bandwidth going in-and-out of memory.
Problem with getting six IoDrives is having six PCI-e slots available that are all x4 or better. However, you can get that same IO rate in IOPS with a pair of high-end 16-port PCI-e SATA/SAS boards, and connect 32 Intel X25-E drives, and likely still have expansion slots left over in your system. That whole setup will burn about 5W per controller, and ~2W per drive under load.
Actually you typically use the outermost, not the innermost. Data rate at the OD is ~2x that of the ID.