Date: Mon, 4 Apr 2011 18:48:25 -0700 From: YongHyeon PYUN <pyunyh@gmail.com> To: Adriel Torres <ladr.torres@gmail.com> Cc: freebsd-net@freebsd.org Subject: Re: Revisit of older discussion on Intel em0 timeouts etc. Message-ID: <20110405014825.GA3420@michelle.cdnetworks.com> In-Reply-To: <BANLkTinWxk2-0gj3NkJUvQ7rCP3bCGuE6Q@mail.gmail.com> References: <BANLkTinWxk2-0gj3NkJUvQ7rCP3bCGuE6Q@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Apr 04, 2011 at 11:21:50AM -0500, Adriel Torres wrote: > Hi, > > I am new to this list and have been reading on the "tcp bug" and "bge > discussions" and wanted to share what I was experiencing with a 8.2 FreeBSD > DNS Server I had recently deployed. Basically, with an em0 interface, link > states would show changes from up to down constantly under very little or no > load. > > After a while, I switched the interface to a dc0-based interface, and for a > while it was fine, then the link states would start to change again. The Would you give me more information on dc(4) link state changes? Can you see abnormal kernel message for dc(4)? I think I fixed a couple of dc(4) issues as well as bus_dma(9) bugs and I don't know how the link state change can happen under load unless it encountered watchdog timeouts or TX underun. It would be even better if you can try latest dc(4) in stable/8 on 8.2-RELEASE. > only other discussion I found where this was talked about in-depth > was the following, but in this case it was occuring for 7.1 stable. > > http://lists.freebsd.org/pipermail/freebsd-net/2010-February/024664.html > > I just wanted to chime in in the event this might be related to any of the > bugs currently on the bug list, and if they are, I would appreciate if > someone could point me to the right discussion/bug. Thank you.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110405014825.GA3420>