SSD power loss report updates

[Alan_F_Intel]

Intel is aware of the customer sightings on Intel SSD 320 Series. If you experience any issue with your Intel SSD, please contact your Intel representative or Intel customer support (via web: http://www.intel.com/ www.intel.com or phone: http://www.intel.com/p/en_US/support/contact/phone www.intel.com/p/en_US/support/contact/phone) . We will provide an update when we have more information.

Alan

Intel's NVM Solutions Group

[idata]

koitsu wrote:

You should be doing backups regardless if you're using an SSD or not.

Not necessarily. I do not keep any irreplaceable data on my boot drives, which are mostly SSDs. So it is a question of whether it is more of a hassle to make regular image backups of my boot drive (X minutes/year), or the hassle to reinstall windows and reinstall/reconfigure my applications weighted by the chance of that happening (Y minutes x AFR). With the X25-M, the annual failure rate seems to be in the 0.5% to 1% range. Assuming 1%, then Y would have to be more than 100 times X to make backing up worth my while. It is not, so I do not back them up.

With the 320, however, it is hard to say. It is a closer decision since the AFR is probably significantly higher with the 320.

[idata]

@ slyphnier

The debate about failure rates is pointless, as nobody knows and nobody has the data (except Intel). So let's not keep bringing it up, yeh?

The reason why people are reacting badly to your posts as you seem to think the burden of responsibility is on the community to fix this bug. It's not. Sure, if you want to collect thousands antedotal reports and try to form some pet theory about the cause, go ahead. What then? Are you going to post the solution to this forum and become hero of the day? I look forward to it.

The ONLY reason to choose the Intel 320 over other drives is the reliability. There are much faster drives out there at the same price. That means the burden on Intel for fixing even minor bugs is much greater than other manufacturers BECAUSE WE ARE PAYING A LOT EXTRA FOR SUPER-RELIABILITY that businesses and certain consumers want.

Even if it just a perception and not a reality, if it is seen that Intel 320s are not so super-reliable, then that is going to hurt sales. I am holding off buying one until it is fixed, how many other people are doing that? How long will we wait? Maybe Intel are big enough not to care, but I am hoping that they are paying the proper attention. A little more communication would be nice, the fact that it is a big company is not an excuse not to communicate with customers. Are you listening, Intel? Wake up!

Regards,

Lloyd

[idata]

Lloyd wrote:

The reason why people are reacting badly to your posts as you seem to think the burden of responsibility is on the community to fix this bug. It's not.

The problem I have with this statement is the use of the word "bug". Let's recap please -- all we know at this point, factually, is this:

1) /message/132261# 132261 A useful post from user carsten indicates the (his) drive starts returning an incorrect maximum capacity value of 1172123568 LBAs (1172123568 * 512 = 600,127,266,816, or around 600GB). I don't know which model of 320-series drive he has, but obviously none of them are 600GB. SSDs don't have an HPA to store this stuff in, so we're not sure where this data comes from (an EEPROM or Flash ROM on the board? Stored in an unused/special area of NAND flash that isn't normally accessible? We don't know!).

2) The issue begins after power is lost to the drive in some way (e.g. during a power outage, someone yanking the cable, suspend/resume, or similar things). At least this is my understanding; I haven't seen reports of someone saying they soft-rebooted their system and found this issue happening.

I really wish I could get my hands on one of these failed drives. None of my 320-series drives have this problem, so I'm both thankful and frustrated at the same time (thankful that I haven't seen it, and frustrated because as someone familiar with low-level storage subsystems I wish I could help more!).

Intel started this specific thread on July 11th to state that they acknowledge the problem -- which is better than nothing -- but no updates have been provided. Only 10 days have gone by since Intel acknowledged they were beginning an investigation! Ten days isn't enough time. So these are your options:

1. Be patient and wait for Intel to provide something conclusive,
2. File an RMA and get a replacement drive (note: remember that Intel Technical Support is not responsible for this issue, nor fixing it -- so please don't give any of the folks on the phone grief over it!),
3. Return the drive to your distributor for a refund. No guarantees they will give you one. If they don't, see option # 2, then sell the replacement on eBay or Amazon or some other place,
4. If you haven't bought one and are excessively worried, then don't purchase one and go with another brand.

There is no piece of hardware that offers "EXTRA SUPER-RELIABILITY". Here's a real-life example in the enterprise world: we just purchased some advertised-as-carrier-grade Ethernet tap devices at my workplace (Fortune 500 company), total cost of the units is around $15K (fiften thousand US dollars) per unit, and we bought 12. A couple days ago we found that there is a major hardware bug with their Ethernet tap hardware where rebooting the product results in all copper modules (only copper, not the fibre modules!) losing link. The Ethernet PHYs are being reset. This should not happen with a tap device (taps are supposed to be 100% non-intrusive). The vendor came back and told us a newer version of their ethernet module will fix this problem, but we have to pay for the upgrade. I've seen this same tactic happen with other enterprise-grade products; it takes months for companies to resolve stuff like this, given limited numbers of engineers, managerial red tape, and so on. So my point is that failures/issues happen no matter what the retail cost of the product is, and consumer vs. "enterprise" grade really means absolutely nothing -- it's a fake/false marketing label.

I myself did not go with an Intel SSD because of "EXTRA SUPER-RELIABILITY", I went with them for other reasons. Will I change vendors? http://www.engadget.com/2011/06/08/corsair-recalls-its-120gb-force-3-ssd-due-to-stability-issues/ With nonsense like this, probably not. But either way we, as consumers, have the ability to choose. There are lots of brands on the market right now. Just remember that nothing is flawless, and to take necessary precautions with that in mind.

[idata]

Some words of wisdom there koitstu, thanks. And true, it has only been ten days - time goes slow when your waiting for news, seemed longer when I was writing that post! Still, a little reassurance like "We have a team of engineers actively working on the problem" and a little more detail would go a long way. For all we know, the effort so far could be a sticky note on somebodies monitor.

Lloyd

[idata]

koitsu,

Just wondering what makes you think the LBA value reported by my drive is incorrect?

What did you expect to see from a 600 GB drive?

My SSD is a 320 Series 600 GB as indicated in my first post in this thread.

Secondly, you've claimed a couple of times now that SSDs don't have an HPA.

To my understanding the HPA commands are at the ATA level, and are not specific to mechanical drives.

You can find quite a few articles on the web describing how to configure an HPA on your SSD.

Even the X25-M datasheet from Intel states that the HPA command set is supported.

See 6.1.5. in http://download.intel.com/newsroom/kits/ssd/pdfs/X25-M_34nm_DataSheet.pdf http://download.intel.com/newsroom/kits/ssd/pdfs/X25-M_34nm_DataSheet.pdf if you care.