hdd-benchmark

How to Run

Set up a client computer (usually using a live CD - I use Arch).

Client

1. Connect to network
2. Get to a command line prompt
3. Start ssh systemctl start sshd.service
4. Set a PW passwd
5. Detemine IP ip address | grep 192

Server / Runner Setup and Config

1. ssh into client ssh root@[IP]
2. install git on client pacman --sync --refresh git
3. clone this repo (use HTTPS) git clone https://github.com/VincentSaelzler/hdd-benchmark.git
4. start screen (so the program runs even if SSH session is disconnected)

Running

Update Parameters

In util.py change is_prod() to return true.

If you forget to do this, you will likely see output that looks like this. It will just go on forever.

before dev sda
before smart FIRST sda
before match: /dev/sda - "in progress, 70% remaining"
in match: /dev/sda
Waiting for /dev/sda extended test to complete. 0.03333333333333333m of estimated 29m
post break: /dev/sda

In main.py ensure the device lists match the disks in your system. At time of writing, they are on lines 45 and 46.

    dev_letters = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h']
    devs = [Dev('sd' + dl, now_unix) for dl in dev_letters]

This example is for 8 drives with the naming convention sda,sdb, ... sdh.

Also edit main.py to include an accurate description of the tests (optional). This is mainly useful if doing lots of tests with different sets of disks and/or test scenarios. At time of writing, this is on lines 14 and 15.

Run the Program

1. Required change directory to root of git repo. Some relative file reference code is not robust.
2. cd hdd-benchmark
3. run the script python3 main.py