Article ID: 122650, created on Aug 14, 2014, last review on Feb 17, 2016

  • Applies to:
  • Virtuozzo 6.0
  • Virtuozzo containers for Windows 4.6


If there are hardware problems on a SSD disk holding pfcache and/or CS journals, read cache, you may observe the following symptoms:

  • High IOWAIT/IOLAT values in "pstorage -c CLUSTER_NAME top" for the local CSes
  • pstorage mount point becomes inaccessible
  • VMs get into PAUSED state


The article provides a script for automated checks of all SSD drives, and explains how to check it manually.

Automated check using a script

To check health of all SSD drives in the system, the attached script can be used.

~# ./ 
direct,sda -- health 95
megaraid,9 -- health 100
cciss,3 -- health 100
aacraid,0,0,3 -- health 80

If the number is close to 1, then the disk is worn out and it needs to be replaced. Replacement instructions are provided in this article.

You can use the displayed identifiers to check details of the disk.

  • direct,NAME

    ~# smartctl -a /dev/NAME
  • megaraid,NUM

    ~# smartctl -a /dev/sdX -d sat+megaraid,NUM
    ~# smartctl -a /dev/sdX -d megaraid,NUM

    Where /dev/sdX is a disk name provided by this RAID. sat+ is needed if the disk is SATA one (not SAS).

  • cciss,NUM

    ~# smartctl -a /dev/sdX -d cciss,NUM

    Where /dev/sdX is a disk provided by RAID.

  • aacraid,N,M,O

    ~# smartctl -a /dev/null -d aacraid,N,M.N

Please check with the hardware vendor how to find out the proper disk using the port number for LSI MegaRAID, HP SmartArray or similar controllers.

NOTE: On servers with AACRAID installed, it is necessary to use smartctl 6.x. You may need to build it from sources:

~# yum install -y subversion automake gcc-c++
~# svn co src/smartmontools-6.4/
~# cd src/smartmontools-6.4
~# ./
~# ./configure 
~# make 
~# cp smartctl /usr/local/bin/

Manual checks

To verify if the SSD disk is healthy, it is necessary to use "smartctl". For Intel SSD disks this can be done using this command:

~# smartctl -a /dev/sdd | egrep 'ID#|Media_Wearout_Indicator|Remaining_Lifetime_Perc|Wear_Leveling_Count'

NOTE: replace "/dev/sdd" with the correct device name in your system.

smartctl parameters may differ greatly, depending on the SSD manufacturer and the underlying driver of the RAID controller.

E.g. for LSI Megaraid with megaraid-sas driver, the following command should be used:

~# smartctl -a -d sat+megaraid,1 /dev/sdd

For SAS disks the following:

~# smartctl -a -d megaraid,1 /dev/sdd

where 1 is the port number of the disk in question.

The new SSD disk should have value '100' for Media_Wearout_Indicator parameter (or any equivalent parameter). If this parameter have the value close to '0' it means that there are problems with SSD disk.

For SAS disks instead of Media_Wearout_Indicator, SS Media used endurance indicator is used. If SS Media used endurance indicator has a value of, for example 2% then SAS drive is healthy. If it is close to 100% then SAS has problems.

Real-life example:

~# smartctl -a /dev/sdd | egrep 'ID#|Media_Wearout_Indicator|Remaining_Lifetime_Perc|Wear_Leveling_Count'

    233 Media_Wearout_Indicator 0x0032   001   001   000    Old_age  Always       -       0

For SAS disks:

~# smartctl -a /dev/sdd | grep 'SS Media used endurance indicator'
    SS Media used endurance indicator: 100%

Search Words


verify ssd

cloud storage




hight iowait

drives showing as failed


c62e8726973f80975db0531f1ed5c6a2 2897d76d56d2010f4e3a28f864d69223 0dd5b9380c7d4884d77587f3eb0fa8ef 6c20476fe6c3408461ce38cbcab6d03b d02f9caf3e11b191a38179103495106f 965b49118115a610e93635d21c5694a8

Email subscription for changes to this article
Save as PDF