Brocade ICX 64XX network switch notes

The Brocade ICX 6430-24 is a 24 port plus four SPF (1gb) ports

The Brocade ICX 6450-224 is a 24 port plus four SPF+ (10gb) ports (albeit, only two 10gb ports are licensed from the factory)

The "p" at the end of the model means POE. The CLI is very close to Cisco's.

Brocade Datasheet

My re-setup notes:

Serial console cable into the switch, hit the reset button, during the boot process tap the "b" key a bunch of times to get to the boot loader.  Once there type: "no password", this disables the password for the next boot.  Then type "boot", and the switch will boot.  Once it boots:

enable

erase startup-config

reload

enable

config terminal

username admin privilege 0 password password

ip dhcp-client enable

crypto-ssl certificate generate

aaa authentication login default local

aaa authentication web-server default local

web-management https

wr mem

To set a static IP:

ip dhcp-client disable

vlan 1

router-interface ve 1

exit

interface ve 1

ip address 192.168.1.55/24

exit

write mem

Further reading: https://fohdeesha.com/docs/index.html

As noted earlier, licencing.......the person in the above link has a way to re-write the serial number, and apply a license key.   FWIW this hardware is well over 10 years old, and Brocade has chosen to ignore this legacy hardware, as it is all End Of Life and End Of Support.

Hard drive e·lec·tric·i·ty and performance

It is common knowledge that SSD's are more energy efficent, but just how much, and does anyone really look at the numbers?

Test Rig:

-Dell Precision T30, Intel G4400 CPU, 8gb DDR4 RAM, Windows 10 installed onto Gen4 NVMe drive. MicroSemi 12gb SAS HBA card.

Test proceedure: all tests were done with SAS card installed, Windows 10 was booted up, and let idle for 3 minutes before taking inital reading.  Atto Disk Bench was run, and observed peak, and average wattage values from a Kill-A-Watt.  These values are somewhat subjective as we are relying on humans.  Not super scientific, but good enough.

A: Test rig with no drives:...................................27w idle.

B: Test rig with no drives, SAS card installed:.....34w idle.

C: Samsung Evo 840 1tb SATA SSD.................37w idle; 43 avg; 60w peak

D: Seagate Barracuda 250gb 7200rpm SATA....38w idle; 47 avg; 53w peak

E: Kioxia 1.9tb Enterprise SAS SSD.................38w idle, 45w avg, 62w peak

F: HGST 10tb 7200rpm SAS............................42w idle; 46w avg, 64w peak

Observations and conclusions:  I was suprised that just having the SAS card in the machine consumes roughly 7 watts alone.  Enterprise SSDs get HOT!  Doing two back to back runs of the benchmark the drive is almost to hot to hold. For the home lab I would suggest getting consumer SSD's to save heat and electricty if one doesn't need that last bit of endurance. So if one needs say 8tb of usable space, and speed isn't the primary issue would one be better off running two SATA/SAS drives in a mirror or four 2tb SSDs or three 4tb SSDs in a RAID5?  Well basic math seems to point that spinning drives would not only be cheaper to buy but to run.  Obviously speed and longer Mean Time Between Failure is longer.