disk issue

T

tincboy

Renowned Member
Apr 13, 2010
466
6
83
Today my server had about 45 minutes with 100 cpu usage,
and after that many of my VMs are not running and while i try to use lvs command to see my lvms I got this error on many of them

/dev/dm-7: read failed after 0 of 4096 at 0: Input/output error
/dev/sdb: read failed after 0 of 4096 at 0: Input/output error

How can I fix this issue?
 
It's a direct attached.
My own's log files show that in that period the IO was not that much to cause this issue, but the CPU was in 100% utilization.
Do you think the high CPU utilization made this situation?
 
tincboy said:
It's a direct attached.
My own's log files show that in that period the IO was not that much to cause this issue, but the CPU was in 100% utilization.
Do you think the high CPU utilization made this situation?
Click to expand...
No!
Are the CPU-Usage high, or "only" the load? If some processes wait for io (due to a defective disk) your load rises very fast (depends on your processes, but 30 is quick reached).

Any info with
Code: 
apt-get install smartmontools

smartctl --all /dev/sdb

Udo
 
thanks for your quick respond,
my logs show's 100% CPU utilization on 8 core CPU with average load of 13
but about 3 hours before HDD issue the load was about 30 for 3 minutes.

would you please explain that the smartmontools is going to help me in which way?
 
Hi,
the smartctl can read the S.M.A.R.T-information of the disk - not all disks support this feature. The disk tracks (read) errors and so on.
This functionality are common on "good" disk like scsi-disks. S.M.A.R.T are created, to bring parts of healty-service also to cheap sata disks.

Udo
 
You must log in or register to reply here.
Top Bottom