3,040 karma · joined January 14, 2018
Personal: office@neil.computer Work: office@usgraphics.com
[1] https://neil.computer/notes/introducing-berkeley-mono/
[2] https://neil.computer/notes/berkeley-mono-december-update/
Update:
Ok, so I used the ':' syntax in the --filename option, I got something that makes sense. FIO created invidual files for each thread. I test numjobs: 4, 8, 16 and it plateaus around 8 jobs. Adding additional files+jobs actually makes things worse.
neil@ubuntu:~/mnt/disk$ ls
dummy0 dummy10 dummy12 dummy14 dummy2 dummy4 dummy6 dummy8
dummy1 dummy11 dummy13 dummy15 dummy3 dummy5 dummy7 dummy9
neil@ubuntu:~/mnt/disk$ sudo fio --name=read_test \
> --filename=/home/neil/mnt/disk/dummy0:/home/neil/mnt/disk/dummy1:/home/neil/mnt/disk/dummy2:/home/neil/mnt/disk/dummy3:/home/neil/mnt/disk/dummy4:/home/neil/mnt/disk/dummy5:/home/neil/mnt/disk/dummy6:/home/neil/mnt/disk/dummy7:/home/neil/mnt/disk/dummy8:/home/neil/mnt/disk/dummy9:/home/neil/mnt/disk/dummy10:/home/neil/mnt/disk/dummy11:/home/neil/mnt/disk/dummy12:/home/neil/mnt/disk/dummy13:/home/neil/mnt/disk/dummy14:/home/neil/mnt/disk/dummy15 \
> --filesize=20G \
> --ioengine=libaio \
> --direct=1 \
> --sync=1 \
> --bs=512k \
> --iodepth=1 \
> --rw=read \
> --numjobs=16 \
> --group_reporting
I get around ~22GB/s with 8 jobs, ~15 GB/s with 16 jobs. Thanks for correcting, I'll post an update to the misleading 157 Gb/s... uff.Updated the article: https://neil.computer/notes/zfs-raidz2/#note-5
See note 4: https://neil.computer/notes/zfs-raidz2/#note-4
The initial file was created with fio rw=randwrite, so the dummy data is random.
I think its even more impressive that ZFS can do this. I verified with ext4 fs on a single NVMe drive and no matter what knobs I turn, I am getting around around 3.5 GB/s.
Edit: See note 2 at the end of the article, I posted the raw results of another 100GB file test.
Edit 2: See note 3. I rebooted the VM with 2GB ram (instead of 64GB) and it matches almost exactly as per your prediction ~ 20GB/s. But, it still doesn't explain the absurd 80GB/s sustained reads of a 100GB file with only 64GB ram.
I’ll say it again, be it hostile - designers aren’t raised and educated in user centric design. MacOS is full of transitions that are over 200ms. iOS is worse. I could go on and on. It’s gotten so bad, it’s appalling really.
[1] https://neil.computer/notes/the-design-of-diskprices-com/
I just fear one day some new intern, engineer, designer or manager is going to join and ruin it. It will be a sad day.
You gotta compare apples to apples.
Here is a kitchen scale I bought. Made in Switzerland, no less. It’s bloody beautiful: https://neil.computer/notes/mettler-toledo-ml3002e-teardown/
Does that mean all consumer grade stuff is bad? No. Leatherman tools and 3M consumer grade stuff is usually great. Nalgene bottles. Does that mean all pro quality stuff is great? No. I’ve bought restaurant quality containers, made in California, no less, but are not BPA-free.
My philosophy is quite tangential to the points you’re making:
- Search for stuff in auxiliary / parallel industries.
- Used surplus stuff is usually below consumer grade prices even if it was slightly worse, you're still equal. 9/10 times it is actually the opposite - 10x the quality, performance and durability of consumer grade crap on Amazon.
- Some cheap crap is actually good. Victorinox $30 knife is excellent value and used by professional chefs. The same cannot be said for an outdoors knife, you want it to be overbuilt. Cheap $30 outdoor knives would not stand a chance. Always buy generic drugs, they're exactly the same.
There are atleast 100 more examples I can write about that were a net positive buys. Just last week I bought U-Line tape dispenser. Same price as consumer grade ones, so why not?
- Orthogonal search usually yields frugal jackpots. No one would ever find any use for the life raft bag in the marine parts store.
Average consumers want stuff that is objectively worse for them. Touchscreens were an excellent idea for a phone, now they want it everywhere in their cars. Overall, with the throw-away culture consumers are addicted to, it’s hard to not blame them. Have you ever dealt with customer service? You might be surprised. We are soooo addicted to wasteful culture. You’re just promoting this and doubling down on it.
Let’s see in last 20 years: Quality went downhill, durability is an after thought, repairability is non existent and consumers want bigger box - literally larger than the item because they choose it based on how good the box looks. If it were any other reason, it’s a waste of packaging and businesses know this. Businesses are not looking to waste money on extra packaging unless it led to better sales. This is a direct evidence of the utter shitshow consumer culture embodies.
Anyways, I fancy things that are engineered with deliberation and specification. Tested and qualified against requirements. Usually military, aviation, marine or medical grade goods are designed and engineered in such a way.
For what it’s worth - I’ve also bought overkill items and they’ve turned out to be a disaster. For instance, for the TPI fan in the example - I had to get an industrial fan controller because it was too damn loud and restaurant grade containers that are not BPA-free.