1 |
On Tuesday, 28 September 2021 12:38:42 BST Rich Freeman wrote: |
2 |
|
3 |
> You keep mentioning USB3, but I think the main factor here is that the |
4 |
> external drive is probably a spinning hard drive (I don't think you |
5 |
> explicitly mentioned this but it seems likely esp with the volume of |
6 |
> data). That math works out to 78MB/s. Hard drive transfer speeds |
7 |
> depend on the drive itself and especially whether there is more than |
8 |
> one IO task to be performed, so I can't be entirely sure, but I'm |
9 |
> guessing that the USB3 interface itself is having almost no adverse |
10 |
> impact on the transfer rate. |
11 |
|
12 |
I'm sure you're right Rich, and yes, this is 2.5" drive. I'm seeing about |
13 |
110MB/s reading from USB feeding into zstd. |
14 |
|
15 |
> The main thing to avoid is doing other sustained read/writes from the |
16 |
> drive at the same time. |
17 |
|
18 |
Quite so. |
19 |
|
20 |
> It looks like you ended up doing the bulk of the compression on an |
21 |
> SSD, and obviously those don't care nearly as much about IOPS. |
22 |
|
23 |
Yes, input from USB and output to SSD. |
24 |
|
25 |
> I've been playing around with lizardfs for bulk storage and found that |
26 |
> USB3 hard drives actually work very well, as long as you're mindful |
27 |
> about what physical ports are on what USB hosts and so on. A USB3 |
28 |
> host can basically handle two hard drives with no loss of performance. |
29 |
> I'm not dealing with a ton of IO though so I can probably stack more |
30 |
> drives with pretty minimal impact unless there is a rebuild (in which |
31 |
> case the gigabit ethernet is probably still the larger bottleneck). |
32 |
> Even a Raspberry Pi 4 has two USB3 hosts, which means you could stack |
33 |
> 4 hard drives on one and get basically the same performance as SATA. |
34 |
> When you couple that with the tendency of manufacturers to charge less |
35 |
> for USB3 drives than SATA drives of the same performance it just |
36 |
> becomes a much simpler solution than messing with HBAs and so on and |
37 |
> limiting yourself to hardware that can actually work with an HBA. |
38 |
|
39 |
-- |
40 |
Regards, |
41 |
Peter. |