What SSD drives does Isilon use in it's nodes

Here is a script that packages up isi_radish -q into something which pull out the manufacturer, model and bay numbers.  This works well since it gets some additional information that is not available in isi_radish -q, namely the manufacturer, for some drives.  I do not pull the drive size but you could modify as necessary to do so.

#!/bin/zsh
typeset -A drive_hash
DRIVE_GCONFIG_FILE=/etc/gconfig/drive_config.gc
REGEXP='.*"(.*)".*'

IFS=$'\n' array_of_lines=($(grep -E "(model|manufacturer).*=.*" "${DRIVE_GCONFIG_FILE}"))
for (( i=1; i <= ${#array_of_lines[@]}; i+=2 )); do
  model=`echo "${array_of_lines[$i]}" | sed -E 's/.*"(.*)".*/\1/'`
  mfg=`echo "${array_of_lines[$i+1]}" | sed -E 's/.*"(.*)".*/\1/'`
  drive_hash["${model}"]="${mfg}"
done

IFS=$'\n' array_of_drives=($(isi_radish -q | sed -E 's/Bay *([0-9]+).*is (.*) FW:.*/\1,\2/'))
for (( i=1; i <= ${#array_of_drives[@]}; i++ )); do
  parts=("${(s/,/)array_of_drives[$i]}")
  bay=${parts[1]}
  mfg=${drive_hash["${parts[2]}"]}
  model=${parts[2]}
  echo "Bay: ${bay}, Manufacturer: ${mfg}, Model: ${model}"
done

Thanks for posting Andrew 

Super!

Now the question I never dared to ask:

Why is it called "radish"...?

-- Peter

RAw DISk Heuristics

{Thanks Kip}

--

Thanks again! -- P.

Andrew,

If you have time, what would the script require to include the field "capacity"?

/etc/gconfig/drive_config.gc

[...]

      # Hitachi 400GB SAS SSD (Sunset Cove)

        hitachi_ssd_10 = {

                model = "HGST HUSMM8040ASS200";

                manufacturer = "Hitachi";

                fwversions = [ "A100" ];

                capacity = 400_gigabyte;

                profile = ssd_sas;

                blocks = 781422768;

[...]

This will give it to you in "marketing" numbers:

#!/bin/zsh

typeset -A drive_hash

typeset -A drive_cap

DRIVE_GCONFIG_FILE=/etc/gconfig/drive_config.gc

REGEXP='.*"(.*)".*'

IFS=$'\n' array_of_lines=($(grep -E "(model|manufacturer|blocks).*=.*" "${DRIVE_GCONFIG_FILE}"))

for (( i=1; i <= ${#array_of_lines[@]}; i+=3 )); do

  model=`echo "${array_of_lines[$i]}" | sed -E 's/.*"(.*)".*/\1/'`

  mfg=`echo "${array_of_lines[$i+1]}" | sed -E 's/.*"(.*)".*/\1/'`

  blocks=`echo "${array_of_lines[$i+2]}" | cut -f3 -d ' '|sed -E 's/;//'`

  drive_hash["${model}"]="${mfg}"

  drive_cap["${model}"]=${blocks}

done

IFS=$'\n' array_of_drives=($(isi_radish -q | sed -E 's/Bay *([0-9]+).*is (.*) FW:.*/\1,\2/'))

for (( i=1; i <= ${#array_of_drives[@]}; i++ )); do

  parts=("${(s/,/)array_of_drives[$i]}")

  bay=${parts[1]}

  mfg=${drive_hash["${parts[2]}"]}

  model=${parts[2]}

  blocks=${drive_cap["${parts[2]}"]}

  ((gb = 1.024 * blocks / 2000000))

  echo "Bay: ${bay}, Manufacturer: ${mfg}, Model: ${model} Capacity: ${gb%.*}"

done

This yields something like this on S210:

# zsh /ifs/data/sarc/drive_info.sh

Bay: 1, Manufacturer: Hitachi, Model: HGST HUSMM1680ASS200 Capacity: 800

Bay: 2, Manufacturer: Hitachi, Model: HITACHI HUC109090CSS600 Capacity: 900

Bay: 3, Manufacturer: Hitachi, Model: HITACHI HUC109090CSS600 Capacity: 900

Bay: 4, Manufacturer: Hitachi, Model: HITACHI HUC109090CSS600 Capacity: 900

This came from:  "Product: S210-2U-Dual-128GB-2x1GE-2x10GE SFP+-21TB-800GB SSD"