I bricked one 750 ssd after about 2 hours of on time. The second replacement ssd died after a few days.
I'm kinda wondering what's happening here.
The 1st drive died before I could really use it and has been replaced. The second was working fine as my root file system and
failed on system power on. I'm waiting to return this device. I'm not getting a replacement.
I'm running a Linux system, 4.2.0 on a p9x79 motherboard. It has 3 pcie 3.0 slots, In both cases the ssd was running fine
and failed to respond on power up.
Are there any Linux s/w tools to query what the problem is?
We are going to check this further. Meanwhile, you can take a look at these links:
http://www.intel.com/support/ssdc/hpssd/ssd-750/sb/CS-035484.htm Intel® SSD 750 Series — Tested Motherboards
http://www.intel.com/support/ssdc/hpssd/ssd-750/sb/CS-035483.htm?wapkw=intel+ssd+750+series Intel® SSD 750 Series — Frequently Asked Questions (FAQ)
Regarding Linux*, you can check the contact methods and recommended forums from the team that designs and maintains the NVME driver for Linux:
Thanks for the response
The Asus p9x79 is not on the list of tested motherboards. But, it has the required pcie 3.0 slots.
I did not enable the UEFI bios. I did not expect to boot from the drive. I also did not want to
add more complications to install.
I was in contact with a very helpful intel software person on the nvme form. I had a problem with the second
card (when it was working) (the first card also.. but I never got a chance to run performance tests) that I could not
achieve the performance benchmarks that were published.
Keith Busch figured it out. I somehow has turned off the MSI-X interrupt support in the kernel. It had been
turned off for some time.
Sadly after Keith found the fix the 750 died. Alas, Keith has no software tools to wake a bricked ssd.
But, it did work and well..till it just stopped.
I bought both cards from Newegg.ca. I can give you the order number and the rma number etc if you want to
track down the first card. When Newegg respond to my last e-mails I can let you know.
We have a few questions in order to understand this issue better.
1- When you mentioned the drive disappeared, did id do it during power on after shut down, or waking up after sleep/hibernation?
2- How did 1st drive die?
3- Was the 1st drive recognized in Bios? 2nd drive?
4- Which Linux distro and version was used?
1. The drive failed during power up. There is no sleep/hibernation on my Linux computer.
When the computer was powered up there was a big delay before the bios message was displayed.
There was also a big delay when the os was booted. The logs:
Sep 07 15:38:45 [kernel] [ 20.821369] nvme 0000:03:00.0: Device not ready; aborting initialisation
Sep 07 15:38:45 [kernel] [ 20.821423] nvme 0000:03:00.0: Device failed to resume
Now, the first drive was never used much. I was working on getting the drive to be the root file system.
The second drive was being used as the root file system and I shut down the system and when it
was powered up again the nvme ssd was not able to be mounted.
I did not see any abnormal shutdown logs.
2. The first drive failed on power on. It was returned.
3. The bios does not support nvme. But, the failed drive (1 and 2) caused a big delay in the boot process.
The bios message (version id etc) took an unusually long time to appear.
4. I'm using a Gentoo distribution. The kernel version was 4.2.0
Additional info: There are 3 pcie 3 slots in the asusp9x79. When the first drive failed I moved
the card to the second slot (it was in the second slot). This did not fix the problem.
When the second drive failed I removed it from the system.
The led's are off after it failed.
There are two other pcie boards in the system: one 680 video card and one pciex1 sound card.
Both the sound and video cards are working fine.
The second card was working fine when it was powered down.
I do not have the logs because they are on the drive. But, here is the hdpam
hdparm -tT --direct /dev/nvme0n1
Timing O_DIRECT cached reads: 3200 MB in 2.00 seconds = 1599.90 MB/sec
Timing O_DIRECT disk reads: 4572 MB in 3.00 seconds = 1523.42 MB/sec
and this might be useful:
I'm sending the defective ssd back to Newegg. If you figure out what the problem is please
let me know. I'm looking at a different ssd. Suggestions would be welcome. Sadly
it cannot be another 750 until the cause of the two previous ssd's is figured out.
But, if the new drive fails I'll let you know.
Thank you for all the information provided. Please know that noted your request in case we hear soemthing similar from other customers. In order for us to properly troubleshoot the drives, we would need to have it available for testing. If you ever face any issue with our drives, please let us know so we can find the best way to assist you.
It is ok, please don't worry about it. We are sure they will contact their suplier to help them with whatever returning procedure they have. It is just for you to know that in case the PoP refuses to replace or even refund, our customer can always start the warranty process with us as long as the issue is not induced.