[Linux-PowerEdge] PE 2950 PERC 5/i slow

Stephen Dowdy sdowdy at ucar.edu
Thu Mar 13 16:07:41 CDT 2014

Dale Dellutri wrote, On 03/13/2014 02:42 PM:
> On Thu, Mar 13, 2014 at 3:36 PM, Jake Smith <jsmith at argotec.com> wrote:
>> We even bought a new 6/I and replaced the 5/i (now it matches the other
>> 2950/6i system) and there was no change in performance.
>> Everything except the CPU and memory chips are identical between them except
>> one hits 140/MBs and the other only 20.
> Just a thought: are the memory chips in the correct slots.  PowerEdge systems
> generally need to have the chips arranged in a certain way to get the full
> performance to memory.  Perhaps the memory bandwidth is the problem?

To check that:

    dmidecode -t memory | sed -ne '/Memory Device/,/Part Number/ { /Size:/h; /^[[:space:]]*Locator:/ {p;x;p}; /Speed:/p}' | paste - - - | tr -s '\t' | expand -t 1,20,50

I'd be surprised if this alone would result in that level of disparity, but you never know!

Also, i would run SMART checks on the disks:

for each disk slot (e.g. 0-8) , and each /dev device (e.g. /dev/sda)

    smartctl -d megaraid,${disknum} -H ${device} | tail -n +3
    smartctl -d megaraid,${disknum} -l error ${device} | tail -n +3
    smartctl -d megaraid,${disknum} -A ${device} | egrep -i '(Reallocated_Sector_Ct|grown defect list|Reported_Uncorrect|Offline_Uncorrectable|Current_Pending_Sector)'
    smartctl -d megaraid,${disknum} -l selftest ${device} | tail -n +3

>> I have two PE 2900's with identical 4x500GB Seagate SATA 7200 RPM drives in

If you see those drive models ending in 340AS, then you have your problem right there ;)
(those drives, and their Nearline SAS brothers, suck,suck,suck, and i'm getting tired of replacing 1 or more a week)

    # /usr/local/LSI-Tools/megacli pdlist a0 | grep Inquiry\ Data
    Inquiry Data:      WD-WCAW3......2WDC WD1003FBYX-18Y7B0                   01.01V03
--> Inquiry Data:             9......0ST31000340NS                                MA0D
(evil, broken drive, note it's the last one to fail in this machine
and be replaced by less evil drives)
    Inquiry Data:             Z......YST1000NM0011                                PA09
    Inquiry Data:             Z......VST1000NM0011                                PA09

or, using smartctl:

    # smartctl -d megaraid,1 -i /dev/sda | grep Model
    Model Family:     Seagate Barracuda ES.2
    Device Model:     ST31000340NS

Might try swapping the controller PCI slot, or check for OS Interrupt sharing?


More information about the Linux-PowerEdge mailing list