| 1 |
Am 17.07.2015 um 18:15 schrieb Daniel Frey: |
| 2 |
> Well, I sure haven't had much luck with SSDs. This will be the third one |
| 3 |
> I've lost. |
| 4 |
> |
| 5 |
> On Wednesday I was watching my mythtv frontend when it hardlocked. Last |
| 6 |
> time this happened the 7-year-old rust recordings drive failed. However, |
| 7 |
> all that checked out and I found out I couldn't ssh in to the frontend |
| 8 |
> to kill mythfrontend. |
| 9 |
> |
| 10 |
> I checked the CPU & RAM by booting via USB and it all checked out. I |
| 11 |
> tried booting the SSD and the kernel panicked. After rebooting again, it |
| 12 |
> started, but every command run ended with a segmentation fault. |
| 13 |
> |
| 14 |
> I decided to try flashing the drive's firmware, and that did so |
| 15 |
> successfully. It booted right away after that with no panic, but the |
| 16 |
> frontend decided that it couldn't find the backend any longer. I found |
| 17 |
> this was not true, I (as root) could ping and connect via mysql using |
| 18 |
> remote credentials. |
| 19 |
> |
| 20 |
> After another twenty minutes of fiddling around, I discovered the setUID |
| 21 |
> root bit on /bin/ping had been removed somehow and this was preventing |
| 22 |
> mythtv from finding its backend. At this point I restored from backup |
| 23 |
> and then I discovered after restoring /bin/ping lost it setuid root bit |
| 24 |
> again. |
| 25 |
> |
| 26 |
> After that I gave up (thinking what else has changed on the disk) and |
| 27 |
> yesterday bought a new SSD, this time a SanDisk model. It was cheap and |
| 28 |
> I hope I don't regret this in the future. So my frontend is once again |
| 29 |
> running. |
| 30 |
> |
| 31 |
> That aside, the drive that failed is a Crucial m4. I have done some |
| 32 |
> searching as how to run diagnostics on an SSD. This drive should still |
| 33 |
> have eight or so months of warranty left. These drive did have a bug if |
| 34 |
> they ran longer than 51xx hours but: |
| 35 |
> |
| 36 |
> 9 Power_On_Hours 0x0032 100 100 001 Old_age Always |
| 37 |
> - 2382 |
| 38 |
> |
| 39 |
> ...there's only 2382 on this drive. It also accesses all media remotely |
| 40 |
> through the LAN. |
| 41 |
> |
| 42 |
> Currently I'm running shred on the affected SSD. I also could run |
| 43 |
> smartctl on the drive. Do other diagnostic tools even work on SSDs? This |
| 44 |
> is where I'm sort of lost, I've not tried diagnostics on them. I usually |
| 45 |
> send them back for warranty, but this time I'm curious. |
| 46 |
> |
| 47 |
> Dan |
| 48 |
> |
| 49 |
> |
| 50 |
|
| 51 |
you know - this does not sound like ssd failure. Most SSDs bomb out by |
| 52 |
just becoming completely unacessible. |
| 53 |
|
| 54 |
dmesg errors? |
| 55 |
|
| 56 |
are you using ecc ram? |
| 57 |
|
| 58 |
if not - maybe, just maybe it is your ram at fault. The stuff the kernel |
| 59 |
sends and the stuff that end on the ssd might not be identical. |
Archives