Date: Tue, 30 Sep 2014 15:19:23 -0300 From: Marcelo Gondim <gondim@bsdinfo.com.br> To: freebsd-net@freebsd.org Subject: Re: Will netmap-ipfw fwd? Message-ID: <542AF42B.60206@bsdinfo.com.br> In-Reply-To: <CA%2BhQ2%2BiH3Ty3M2A=Whfr2SY2Z_WGhr0NezVJ_jcYR-5C7su4fA@mail.gmail.com> References: <CAEqdE_7wiquoeC0P2kwLGnYHqhOat5Bjq8yvebE8U9TPufGsag@mail.gmail.com> <CA%2BhQ2%2BiH3Ty3M2A=Whfr2SY2Z_WGhr0NezVJ_jcYR-5C7su4fA@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 30/09/2014 13:49, Luigi Rizzo wrote:
> Should work.
> Please try the latest version from code.google.com/p/netmap-ipfw/
>
> Cheers
> Luigi
Hi Luigi,
The netmap-ipfw be included in the FreeBSD 10.1 final?
Cheers,
Gondim
>
> On Tuesday, September 30, 2014, Eduardo Meyer <dudu.meyer@gmail.com> wrote:
>
>> I have a problem, where I need to fwd a high rate of pps, and I dont have
>> enough CPU. It's around 900Kpps, so I would like to know if ipfw userland
>> version with netmap support will do fwd?
>>
>> Here are my current rules:
>>
>> 00100 fwd 10.1.2.1 tag tcp from table(100) to any dst-port 80,1024-65535 in
>> { via lagg0 or via vlan1010 }
>>
>> 00200 prob 0.500000 fwd 10.1.2.2 tcp from any 80,1024-65535 to table(100)
>> in { via igb6 or via igb7 }
>> 00300 fwd 10.1.2.3 tcp from any 80,1024-65535 to table(100) in { via igb6
>> or via igb7 }
>>
>> With those rules, my CPU interrupt rate raises from 30% to 80%.
>>
>> If netmap/ipfw has the ability to fwd should I expect a better interrupt
>> rate and lower load?
>>
>> Sorry to ask before actually trying but this is a production environment
>> and I don't have enough room to test, I would like to read opinions before
>> finding out a nightly window to do the changes and tests.
>>
>> Thank you.
>>
>> --
>> ===========
>> Eduardo Meyer
>> pessoal: dudu.meyer@gmail.com <javascript:;>
>> profissional: ddm.farmaciap@saude.gov.br <javascript:;>
>> _______________________________________________
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?542AF42B.60206>