Date: Mon, 28 Jan 2013 09:15:46 -0800 From: Vijay Singh <vijju.singh@gmail.com> To: Luigi Rizzo <rizzo@iet.unipi.it> Cc: net@freebsd.org Subject: Re: ixgbe & msi/x Message-ID: <CALCNsJSK3wO45YE4TaYF=LWcab7iECfdo0FArvLbaAbPT6U9Yg@mail.gmail.com> In-Reply-To: <20130128070439.GB85353@onelab2.iet.unipi.it> References: <CALCNsJRdwKKB5DwRAFmenXkY8u8bmBh7QTjyrjqf_Q1HDw%2B3QQ@mail.gmail.com> <20130128070439.GB85353@onelab2.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
> just curious, is this happening under behyve or also native, > and is it always occurring or it is occasional ? Native, and it happens when the pps rate is high, even if the aggregate bandwidth is low. > I am asking because with netmap when i tried to exploit interrupt > mitigation (strictly processing incoming traffic only on rx > interrupts) i noticed packet drops even at relatively low rates, > which made me suspect that interrupts were either lost or heavily > delayed. I am running whatever is in the driver (version 2.4.5) by default. Since msi/x isnt enabled by default, I have enabled that. The same test, sustains a *much* higher pps load with legacy interrupts, so I think that the msi/x interrupt setup is missing something. -vijay
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CALCNsJSK3wO45YE4TaYF=LWcab7iECfdo0FArvLbaAbPT6U9Yg>