If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
|
Thread Tools | Rate Thread | Display Modes |
#16
|
|||
|
|||
Strange new drive
In article , Lucifer
wrote: Enterprise does not use SSD because they are unreliable. nonsense. enterprise uses ssds because not only are they more reliable but they're also much, much faster. |
Ads |
#17
|
|||
|
|||
Strange new drive
On Sun, 16 Jun 2019 01:47:48 -0500, VanguardLH wrote:
Lucifer wrote: VanguardLH wrote: Consumer-grade SSDs have a small overprovisioning than do SSDs in an enterprise configuration. Longevity and reliability are more important for enterprise usage. Enterprise does not use SSD because they are unreliable. Wrong again. While the SSD already has some overprovisioning, it gets or should get increased in an enterprise scenario. SSDs wear out quickly with high usage. For speed use SAS hard drives. |
#18
|
|||
|
|||
Strange new drive
In article , Lucifer
wrote: Enterprise does not use SSD because they are unreliable. Wrong again. While the SSD already has some overprovisioning, it gets or should get increased in an enterprise scenario. SSDs wear out quickly with high usage. hard drives wear out more quickly, nor matter what usage. For speed use SAS hard drives. nope. for speed, use ssd. for the most capacity per dollar, use sata drives. |
#19
|
|||
|
|||
Strange new drive
Lucifer wrote:
On Sun, 16 Jun 2019 01:47:48 -0500, VanguardLH wrote: Lucifer wrote: VanguardLH wrote: Consumer-grade SSDs have a small overprovisioning than do SSDs in an enterprise configuration. Longevity and reliability are more important for enterprise usage. Enterprise does not use SSD because they are unreliable. Wrong again. While the SSD already has some overprovisioning, it gets or should get increased in an enterprise scenario. SSDs wear out quickly with high usage. For speed use SAS hard drives. Yep, deny what the experts say. |
#20
|
|||
|
|||
Strange new drive
On Sun, 16 Jun 2019 14:51:37 -0500, VanguardLH wrote:
Lucifer wrote: On Sun, 16 Jun 2019 01:47:48 -0500, VanguardLH wrote: Lucifer wrote: VanguardLH wrote: Consumer-grade SSDs have a small overprovisioning than do SSDs in an enterprise configuration. Longevity and reliability are more important for enterprise usage. Enterprise does not use SSD because they are unreliable. Wrong again. While the SSD already has some overprovisioning, it gets or should get increased in an enterprise scenario. SSDs wear out quickly with high usage. For speed use SAS hard drives. Yep, deny what the experts say. I read where somebody is using very high temperatures to recover worn out SSDS. |
#21
|
|||
|
|||
Strange new drive
Lucifer wrote:
I read where somebody is using very high temperatures to recover worn out SSDS. Yes, I remember hearing about that, too. Something to do with healing the oxide stress of the junctions. Probably not something that end users could do without the expertise of silicon doping and the proper equipment. High temperatures encountered in application (i.e., in a computer setup) are harmful but the healing temperatures have to be much higher, highly localized and, I suspect, require popping apart the casing to expose the wafer and probably incorporate an scope to monitor the process of reflow. I haven't seen anything about this lately, but then I haven't been looking (I wasn't looking before and just ran across such reports of high-temperature healing). Seemed more like an experiment to prove capability than of economical restoration. There have been some studies of incorporating high-temperature healing within the chip itself; i.e., to produce self-healing SSDs. This would be far more economically feasible than having labs do the work. Recovery is possible in labs for HDDs, but that is an expensive repair that most end users are unwilling to pay, but companies may employ the service for business-critical data (but implies the company didn't employ sufficient and multiple backup strategies). Exploiting Heat-Accelerated Flash Memory Wear-Out Recovery to Enable Self-Healing SSDs https://pdfs.semanticscholar.org/66f...3cbba017c8.pdf (not dated but the latest cited reference is 2009) That seems an old article, and I've not heard of any self-healing SSDs at all. The additional circuitry would likely raise the cost of the flash modules by a huge increase putting it outside the consumer market until something like 6-10 years have lapsed to lower the manufacturing cost to an effective price point acceptable to the consumer market. It would have to be the near the same cost, perhaps a bit more expensive, to buy a self-healing SSD than to buy a new one. Basically the added cost would be like buying insurance on the part. End users often don't monitor temperatures of their drives whether HDDs or SSDs. While the BIOS/UEFI has temperature thresholds for the CPU, they don't for the GPU or drives. I've also found drive monitoring software that relies on SMART attributes (all of them do) to be guessing at the longevity of HDDs. For example, they may see a big jump of 200 sectors getting remapped in one event (between their SMART polls) and assume the drive just must be going bad. Yet SMART does not report the size of the G-list: the spare sector pool. 200 remapped sectors looks bad by itself but only if the G-list were, say, 500 sectors. Not as bad if the G-list were 2000 sectors, and trivial with a G-list of 10,000 spare spectors. A one-time jump of 200 sectors isn't catastrophic, but repeated such jumps would indicate accelerated progressive failures. The drive monitors just go by the size of the remapped sectors per event without any knowledge of what is the G-list size which they cannot determine from the SMART attributes. SSDs, on the other had, do have a SMART attribute to indicate remaining lifespan based on the write volume. High temperatures (not those for doping or reflow of SSDs, but just higher than the operating temperature range for the device) can be encountered in the devices used by end users. Users expect data to be permanent yet they mistreat their devices, like leaving them inside a car with the windows rolled up on a hot summer day. It's illegal to leave babies in that scenario, but users do it to their devices. They shorten the life of their recordable discs (that use dye sublimation instead of burned in pits) as the data fades in heat. SSDs can lose their data, too, when exposed to higher temperature than their operating range when not powered on. https://www.extremetech.com/computin...-without-power The chart in the above article claims a powered off SSD has about a 1-week retention at 55C (131F). Yeah, users won't be working in scalding temperatures, but they might store their devices with SSDs that end up encountering those temperatures. A car in direct sunlight on a hot summer day with temperatures of 70 to 100 F can reach inside temperatures of 130 to 172 F. If a data center should lose its air conditioning, wonder what the temperatures inside will rise to. Enterprise-class SSDs have a higher sized G-list but that's for remapping the bad sectors, not to offset data loss due to storage at high temperatures. Is There Really a Difference Between Enterprise and Consumer Solid State Drives? https://www.itprotoday.com/high-spee...d-state-drives Enterprise versus Client SSD https://www.kingston.com/us/ssd/ente...sus_client_ssd For the above mentioned self-healing SSDs to work, they may have to be SLC types, like the enterprise-class SSDs. Rare few end users buy SLC SSDs. Too pricey for their pockets. Price is already and still a deciding factor of where to deploy SSDs in an enterprise setup, like a data center, and self-healing SSDs, if ever available, would up the price curve again. |
#22
|
|||
|
|||
Strange new drive
In article , Lucifer
wrote: Enterprise does not use SSD because they are unreliable. Wrong again. While the SSD already has some overprovisioning, it gets or should get increased in an enterprise scenario. SSDs wear out quickly with high usage. For speed use SAS hard drives. Yep, deny what the experts say. I read where somebody is using very high temperatures to recover worn out SSDS. so what? there are people who recover worn out hard drives too. lots of them. |
#23
|
|||
|
|||
Strange new drive
nospam posted this via
: In article , Lucifer wrote: Enterprise does not use SSD because they are unreliable. Wrong again. While the SSD already has some overprovisioning, it gets or should get increased in an enterprise scenario. SSDs wear out quickly with high usage. For speed use SAS hard drives. Yep, deny what the experts say. I read where somebody is using very high temperatures to recover worn out SSDS. [Dogma and tripe snipped. Nothing left.] You're such a ****y little girl. -- I AM Bucky Breeder, (*(^; Resolve conflicts the American way : Rock - Paper - Scissors - Aircraft Carrier Battle Group .... and I approve this message! |
|
Thread Tools | |
Display Modes | Rate This Thread |
|
|