FreeBSD as an embedded platform

Gerry Weaver gerryw at compvia.com
Sat Oct 3 08:56:41 UTC 2015


Hi,

The main test was a point-to-point vpn tunnel. I was looking at bandwidth and a load balancing algorithm. I do happen to have some additional hardware, so I'll install 10.2. on another pair of appliances and test my code on both versions. Hopefully, that will give me something fairly specific to come back with. 

Thanks,
-G


-----Original Message-----
From: Adrian Chadd [mailto:adrian.chadd at gmail.com] 
Sent: Tuesday, September 29, 2015 2:10 PM
To: Gerry Weaver <gerryw at compvia.com>
Cc: freebsd-hackers at freebsd.org
Subject: Re: FreeBSD as an embedded platform

On 28 September 2015 at 22:41, Gerry Weaver <gerryw at compvia.com> wrote:
> Hi,
>
> It wasn't anything exhaustive. I just ran some benchmarks on some of my code. It is mainly oriented around networking and crypto. I also noticed a difference in general system responsiveness. We have a couple of servers that we use for virtualization storage that we upgraded to 10.1 for some infiniband testing. We ended up rolling those back to 9.3 as well. I/We have not attempted to rebuild with gcc yet as you suggested. Using a stock build is preferred where possible.
>
> Disclaimer: I didn't intend this post to be a complaint. I fully understand the reasoning behind the tool chain switch. I'm just wondering what other folks are doing and what challenges, if any, they have encountered.

Well, a lot has changed between 9 and 10. I'd rather we figure out what is not working well for you and fix it so you don't have to roll back. Otherwise you're going to be stuck on 9.3 when it goes EOL and then be forced to suffer the pain. :)

So, what were the benchmarks, and how'd htey compare?


-adrian


More information about the freebsd-hackers mailing list