By Ali Hassan, Editor, UAE · Updated 11 November 2026

How to Export Server & Switch Hardware Specs for Quotes

Nobody wants to spend a morning at a rack typing inventory by hand. You don’t have to. Before you power anything down, five minutes of copy-paste commands can capture the CPU, memory, drive and module details a buyer needs. This guide shows how to export server and switch hardware specs for buyback on every common platform.

Quick answer: Run the platform’s own inventory command, save the output as text, and send it with serial numbers: show inventory on Cisco, racadm hwinventory on Dell, the iLO Redfish API on HPE, and dmidecode plus lshw on Linux.

The cheat sheet

PlatformCommandWhat it captures
Cisco IOS XE / NX-OSshow inventoryChassis, PSUs, fans, modules: PID and serial
Ciscoshow versionModel, software release, uptime
Dell iDRACracadm hwinventoryCPUs, DIMMs, NICs, controllers
Dell iDRACracadm getsysinfoModel and service tag
Dell iDRACracadm raid get pdisks -oPhysical drives
HPE iLOGET /redfish/v1/Systems/1/Model, serial, summary
Linuxdmidecode -t processor -t memoryCPU model, DIMM size, speed, part number
Linuxlscpu, lsblkCPU stepping, drive models

Cisco: show inventory to CSV

Turn off paging, then capture the output through your terminal’s session log:

terminal length 0
show version
show inventory
show switch

On Nexus, add show module. To get a CSV, run this small script on the saved text (test it on one device first):

python

import re, sys t = open(sys.argv[1]).read() for n, d, p, v, s in re.findall(r'NAME: "(.*?)", DESCR: "(.*?)"\s*PID: (\S+)\s*,\s*VID: (\S*)\s*,\s*SN: (\S*)', t): print(f'"{n}","{d}",{p},{v},{s}')

Dell and HPE: ask the BMC, not the OS

The BMC reports hardware even when the operating system is dead, which is common with retired servers. For Dell:

racadm --nocertwarn -r IDRAC_IP -u USER -p PASS hwinventory > dell-hwinv.txt
racadm --nocertwarn -r IDRAC_IP -u USER -p PASS getsysinfo > dell-sysinfo.txt
racadm --nocertwarn -r IDRAC_IP -u USER -p PASS raid get pdisks -o > dell-disks.txt

For HPE, use iLO’s Redfish API:

curl -sk -u USER:PASS https://ILO_IP/redfish/v1/Systems/1/ > hpe-system.json
curl -sk -u USER:PASS https://ILO_IP/redfish/v1/Systems/1/Processors/ > hpe-cpu.json
curl -sk -u USER:PASS https://ILO_IP/redfish/v1/Systems/1/Memory/ > hpe-memory.json

Collection URLs return links to each member (every CPU, every DIMM). Follow those links to get the detail. Our HPE server selling guide explains what buyers look for.

Linux: dmidecode and lshw

Run as root. This captures system, CPU, memory, drives and storage or network controllers in one file:

sudo bash -c 'dmidecode -t system -t processor -t memory; lscpu; lsblk -o NAME,MODEL,SIZE,SERIAL; lspci | grep -Ei "raid|ethernet"' > $(hostname)-specs.txt
sudo lshw -short >> $(hostname)-specs.txt

lscpu shows the CPU stepping. See the dmidecode man page for more type options.

Handle the outputs safely

  • Use a read-only BMC account and delete it afterwards.
  • Strip passwords, IP addresses and hostnames before emailing.
  • Clear shell history if you typed credentials.
  • Keep serial numbers. Buyers need them.

What to send with the files

Send one text file per device, a rack location list, condition notes (cracked bezel, missing caddy), and a photo of each serial label. Then wipe or destroy data after exporting, never before. Our data destruction process is aligned with NIST SP 800-88.

Ready for a quote?

Send your files to our IT equipment buyback team. For switches, see our Cisco selling guide.