>On Fri, 1 Mar 2024 09:24:15 +0100
>Marco Feichtinger wrote:
>
> The mirrored disks are 480GB ssds.
>
> The copying with dd(1) took ~16 hours.
> Is that normal?
>
> -marco
Hi, a long-time lurker here. If 480 GiB are transferred over 16 hours,
the implied transfer rate is 70 Mbps. If 480 GB
>On Fri, 1 Mar 2024 10:08:25 +0100
>Marco Feichtinger wrote:
>
> > and the computer isn't a SBC bitty box, the transfer rate is weirdly
> > low.
>
> Well, both disk are on the same machine.
> It's a Supermicro X7SPA-H-D525 board.
>
> -marco
Well, that's not a bitty box. Wish I bought something
I setup a mirror, according to the instructions on the 9legacy Doc page.
Booted into fossil from disk3, installed fossil on disk1 and copied the data to
disk2 with dd(1).
The mirrored disks are 480GB ssds.
The copying with dd(1) took ~16 hours.
Is that normal?
-marco
> and the computer isn't a SBC bitty box, the transfer rate is weirdly
> low.
Well, both disk are on the same machine.
It's a Supermicro X7SPA-H-D525 board.
-marco
--
9fans: 9fans
Permalink:
Increasing the dd block size (-bs 1024k or as big as the man pages
allow) could make a big difference.
On 3/1/24, Aleksandar Kuktin wrote:
>>On Fri, 1 Mar 2024 10:08:25 +0100
>>Marco Feichtinger wrote:
>>
>> > and the computer isn't a SBC bitty box, the transfer rate is weirdly
>> > low.
>>
>>
A larger block size as lucio says, and also try two dd's with a pipe between
them,
one reading and one writing. dd(1) is single threaded but you have two
asynchronous physical devices.
I have had good success copying sd-cards using fcp - rsc's multithreaded cp,
already in the distribution,
01.03.2024 11:38:06 Steve simon :
> A larger block size as lucio says, and also try two dd's with a pipe between
> them,
> one reading and one writing. dd(1) is single threaded but you have two
> asynchronous physical devices.
You can probably even pipe it through some compressor, but I doubt
