You are not logged in.
Hi,
So since a few days, invariably after a little bit of time (generally around one hour), applications begin to be unresponsive, and I can glimpse in the terminal for example some input/output errors.
I tried looking at the journal using
journalctl -b -1
but I never do find any kind of error.
So I don't really know how to diagnosticate this problem.
I thought it could be the SSD failing, so I run an extended test using SMART. Everything seems fine.
I fscked all the partitions (ext4) forcing the verification, everything seems fine.
Do you have an idea of what could cause that ? Or why I can't find the corresponding errors in the journal ?
Thanks for your help,
Sathors
Offline
Would it be possible to post the error messages you are seeing in the terminal?
Offline
Well this time it seems it hasn't lock up. But I have to copy down the log, because previous time there wasn't any way to get it after a reboot.
It seems all the problem begins like that :
# journalctl
udisks-daemon[305]: **** Refreshing ATA SMART data for /sys/devices/pci0000:00/0000:00:1f.2/ata3/host2/target2:0 ... (not wrapped, can't see it)
ata3.00: exception emask 0.0 SAct 0x0 SErr 0x0 action 0x6 frozen
ata3.00: failed command: WRITE DMA
ata3.00: cmd ca/00:40:28:c5:77/00:00:00:00:00/ec tag 0 dma 32768 out
res 40/00:00:00:4f:c2/00:00:00:00:00/00 Emask 0x4 (timeout)
ata3.00: status: { DRDY }
ata3.00: hard resetting link
ata3.01: hard resetting link
ata3.00: link is slow to respond, please be patient (ready=0)
ata3.00: SRST failed (errno=-16)
ata3.00: hard resetting link
ata3.01: hard resetting link
ata3.00: link is slow to respond, please be patient (ready=0)
ata3.00: SRST failed (errno=-16)
ata3.00: hard resetting link
ata3.01: hard resetting link
ata3.00: link is slow to respond, please be patient (ready=0)
ata3.00: SRST failed (errno=-16)
ata3.00: limiting SATA link speed to 1.5Gbps
ata3.01: limiting SATA link speed to 1.5Gbps
ata3.00: hard resetting link
ata3.01: hard resetting link
ata3.00: SRST failed (errno=-16)
ata3.00: reset failed, giving up
ata3.00: disabled
ata3.00: device reported invalid CHS sector 0
ata3: EH complete
sd 2:0:0:0: [sda] Unhandled error code
sd 2:0:0:0: [sda]
Result: hostbyte=0x04 driverbyte=0x00
sd 2:0:0:0: [sda] CDB:
cdb[0]=0x2a: 2a 00 0c 77 c5 28 00 00 40 00
end_request: I/O error, dev sda, sector 209175848
EXT4-fs warning (device sda5) : ext4_end_bio:317: I/O error writing to inode 1197549
Buffer I/O error on device sda6, logical block 1581858And the lines since unhandled error code goes on and on all the time. All of this happened in less than 2 secs.
Now an ls gives me :
ls: reading directory .: Input/output errorI cannot use any function of my windows manager awesome, which gives me a
/usr/share/awesome/lib/awful/prompt.lua.53: input/output error
I can open an other terminal, but trying to run xterm at the prompt gives me :
bash: /usr/bin/xterm: Input/output errorDo you have any idea what could cause all of that ? Would it be the SSD faulty, even though it has less than one year and that the SMART extended test succeeds ?
Thanks for your input
Offline
I had this exact issue just before my SSD died (including it passing all the SMART tests).
It's worth testing with a different cable and SATA port in case one of them is faulty.
Offline
purely for my own curiosity... what kind of SSD is this?
Offline
Well if it's going to die that's not exactly a good nows...
For your information, it's a 128go crucial m4.
Anybody that could confirm/infirm the nearing death ?
Thanks
Last edited by Sathors (2014-04-01 22:01:20)
Offline
Most SSDs have warranties from two to five years. So maybe your SSD is still within that range and might be able to be RMA'ed
Offline
Then I'll begin cloning my partitions.
Are you aware of any mean of checking whether the ssd is really going to die ? Something more thorough or more specialized than SMART ?
Thanks
Offline
Sometimes the vendor has a special tool that can check for you, but I am not sure about Crucial/Micron.
When SMART works and indicates a drive going bad, it is a really great thing, but unfortunately it isn't accurate quite as often as one would expect.
Offline
Well well well.
I had done a backup of my home partition a few days ago (the problem had already started) on an external drive using rsync.
I did have some errors on a few unimportant files, saying me multibyte something.
Now when I want to access this backup (for example a 'du -sh backup'), or if I want to copy this backup to an other ssd using rsync, I get a lot of these errors :
rsync -a backup new_home
Buffer I/O error on device sdb1, logical block 1720889078
end_request:critical medium error, dev sdb, sector 172343908(note : /dev/sdb1 is the backup drive)
and sometimes appears the error :
rsync: read errors mapping "backup/somefile": Input/Output error (5)Also I cloned the root partition using dd if=/dev/sda of=/dev/sdc bs=1M comps=noerror,notrunc,sync, and it seems everything went fine without errors.
Do you think I'll still get Input/Output errors although dd didn't tell me nothing ?
Do you think it'll be better to copy using dd or clonezilla, or a filesystem-aware tool like rsync ?
Thanks for the help folks.
Last edited by Sathors (2014-04-02 15:43:57)
Offline