Help with errors with ssd on dell latitude 7270

4573746562616e · 2017-11-25 17:16:17

Hello, first of all thanks for your time. I'm having some issues with my becoming unresponsive and hanging my arch install.

This started happening ~2 of weeks ago, and I regulary update the whole system (pacman -Syu).

I've tried to restore windows, testing the drive, doing the system diagnostics from Dell, memtest, verifying S.M.A.R.T. values and everything looks ok. On windows it doesn't freeze, not even one error on a full install. Running exactly the same system (rsync) in a sdcard works perfectly too.

This is the first error I get, on this error the system continues to work:

https://ptpb.pw/S4IM
https://pastebin.com/u3Mqsgx3

After a while I get this errors:

https://ptpb.pw/GFAH
https://pastebin.com/25h5J1CF

And at that point I was lucky I managed to save the errors to a file elsewhere.

I've some pictures of other random errors I get afterwards but it's mostly ext4 failing because it can't access the drive, I can upload them if necessary.

My smartctl -x output

https://ptpb.pw/4hQf
https://pastebin.com/N2EjRy7q

uname -a
Linux [host] 4.13.12-1-ARCH #1 SMP PREEMPT Wed Nov 8 11:54:06 CET 2017 x86_64 GNU/Linux

I can't reproduce this consistently yet. It just happens at some point. So far I understand it has to be related to the sata controller or some controler of some sort as the drive itself seems to be working fine, or at least all tests report it is working fine.

Any ideas on how to proceed?

I've removed and re-seated the ssd too and the issue stills randomly happens.

I've added this parameter to the kernel wrongly thinking my drive was nvme (and because the errors were similar): nvme_core.default_ps_max_latency_us=5500 anyway didn't seem to have effect at all.

Last edited by 4573746562616e (2017-11-25 17:28:12)

Slithery · 2017-11-25 17:19:17

Welcome to the forums 4573746562616e

Any chance you could post your pastes to a different site?
https://wiki.archlinux.org/index.php/Li … in_clients

4573746562616e · 2017-11-25 17:27:32

Done, thanks for the tip.

cfr · 2017-11-25 22:12:57

4573746562616e wrote:

I can't reproduce this consistently yet. It just happens at some point. So far I understand it has to be related to the sata controller or some controler of some sort as the drive itself seems to be working fine, or at least all tests report it is working fine.

Except the ones which don't?

# 4  Extended offline    Interrupted (host reset)      90%       775         -

Backup.

4573746562616e · 2017-11-25 22:32:38

Hello cfr, I already have a sd card with the whole system and in addition a backup of all personal data. Thanks.

I forgot to mention, that S.M.A.R.T. self-test was canceled by myself by mistake, then I ran it to completion and worked.

If I use the hard drive intensively it tends to fail more often, but only on Linux, when testing with windows I actually did the full installation (I don't use windows, but needed to use the same testing setup that dell warranty would use) and ran stress tests and the smart tools (there is when I canceled it by mistake) and had no errors or problems at all.

I can't say that it "won't" fail on windows as I didn't use it for extended periods of time more than the 8 hours of intentional testing. At that point I was thinking maybe the filesystem of my installation was corrupted due some times the system ran out of battery. So I copied again the system from the sd-card with rsync to the ssd, reconfigured grub and continue working only for it to fail again ~3 hours later.

At the moment I'm trying to find ways to make it fail consistently without success, but it does endup failing one or twice a day now. My plan is to run the system from the sdcard and do some i/o intensive stuff on the ssd that way the system shouldn't die on me when the ssd/controller/driver fails and I can get more information or see if there is a way to force the kernel to try connect the drive again??

I am no expert on linux kernel/driver stuff but for what I grasp the drive stops responding or responds badly and then the kernel decides to disconnect it and ignore it. After that if I restart the system it might just work for the next 4-10 hours or maybe fails immediately.

What I really want to be sure is that if I do buy another SSD it the error will be gone, all backups are done so any tip/help/guide to actually understand which component is failing and why would be greatly appreciated.

Last edited by 4573746562616e (2017-11-25 22:34:09)

cfr · 2017-11-26 14:50:24

4573746562616e wrote:

Hello cfr, I already have a sd card with the whole system and in addition a backup of all personal data. Thanks.
I forgot to mention, that S.M.A.R.T. self-test was canceled by myself by mistake, then I ran it to completion and worked.

Ah. That's different. Just I recently had a drive die and that exact complaint was the first symptom. Re-running the test succeeded and I hoped it was a blip. However, a couple of weeks later, it obviously wasn't.

Firmware up to date? Have you searched for bugs/issues specific to this drive and Linux? (Or this machine and Linux regarding this drive, since it could be the interface rather than the drive itself, I guess.)

There seems to be an upgrade for your SSD's firmware: http://www.dell.com/support/home/uk/en/ … erId=J67P6. I can't get the manufacturer's site up at the moment, but you may find you can also get the upgrade from them. Dell's upgrade requires Windows, which may or may not be a problem, but search results suggests the OEM lists Linux as supported, too, so they should provide a suitable upgrade route.

Last edited by cfr (2017-11-26 15:01:17)

4573746562616e · 2017-11-27 19:59:45

cfr wrote:

...
Firmware up to date? Have you searched for bugs/issues specific to this drive and Linux? (Or this machine and Linux regarding this drive, since it could be the interface rather than the drive itself, I guess.)
There seems to be an upgrade for your SSD's firmware: http://www.dell.com/support/home/uk/en/ … erId=J67P6. I can't get the manufacturer's site up at the moment, but you may find you can also get the upgrade from them. Dell's upgrade requires Windows, which may or may not be a problem, but search results suggests the OEM lists Linux as supported, too, so they should provide a suitable upgrade route.

Well... I took the time to install windows again (zZzZz) update the SSD's firmware (had to do it manually, the Dell update tool didn't recognize the firmware outdated) and restored my arch ~ 8 hours ago. For now no errors occurred; If it doesn't fail in the next 48 hours I will come back to hopefully close this as fixed and update the wiki page (https://wiki.archlinux.org/index.php/De … tude_E7270) with the tip.

Thank you very much for your help.

Last edited by 4573746562616e (2017-11-27 20:00:19)

4573746562616e · 2017-11-28 15:50:52

Bad news... at least for me. Today when I woke up it was dead again. I'm running out of ideas on what to do.

It seems to fail randomly not related on the load. What was strange this time is that after the hard-reset It failed once again immediately after restarting when I was opening the browser to post this.

My new idea is that due the temperature reported via smart is always the same I suspect there is no thermal control at all on this drive (or it's failing).

If I can't find anything else I will try installing a different kernel version before buying a new drive to test... I would have bought one by now but I'm in Argentina at the moment and the local prices for m2 ssd drives are crazy high (200% amazon prices the cheapest).

4573746562616e · 2017-11-30 17:01:57

Well, I've been living on a sdcard for the past 48hs or so, with the drive mounted. For the first 6-8 hours I did stress the drive as much as I could (hashed the whole contents of the drive, written a lot of files with random data, fsck, etc). Now about 10' ago I was going to migrate back to the drive and it failed while trying a directory listing.
dmesg
https://ptpb.pw/Uthb
journal (minus wireless/firewall/apps)
https://ptpb.pw/f-aN

Linux bucefalo 4.13.12-1-ARCH #1 SMP PREEMPT Wed Nov 8 11:54:06 CET 2017 x86_64 GNU/Linux

Maybe it was when the drive tried to go to sleep due to inactivity for a long period of time? How can I test this manually?

EDIT: Well, I've moved to linux-lts kernel and now I'm waiting for it (not) to fail again. If it doesn't fail does this qualify as a kernel bug?

Last edited by 4573746562616e (2017-12-01 19:44:24)

Arch Linux

#1 2017-11-25 17:16:17

Help with errors with ssd on dell latitude 7270

#2 2017-11-25 17:19:17

Re: Help with errors with ssd on dell latitude 7270

#3 2017-11-25 17:27:32

Re: Help with errors with ssd on dell latitude 7270

#4 2017-11-25 22:12:57

Re: Help with errors with ssd on dell latitude 7270

#5 2017-11-25 22:32:38

Re: Help with errors with ssd on dell latitude 7270

#6 2017-11-26 14:50:24

Re: Help with errors with ssd on dell latitude 7270

#7 2017-11-27 19:59:45

Re: Help with errors with ssd on dell latitude 7270

#8 2017-11-28 15:50:52

Re: Help with errors with ssd on dell latitude 7270

#9 2017-11-30 17:01:57

Re: Help with errors with ssd on dell latitude 7270

Board footer