Date: Thu, 20 Feb 2003 11:08:26 -0600 From: Dan Nelson <dnelson@allantgroup.com> To: Will Saxon <WillS@housing.ufl.edu> Cc: questions@FreeBSD.ORG Subject: Re: network tuning Message-ID: <20030220170826.GO13096@dan.emsphone.com> In-Reply-To: <0E972CEE334BFE4291CD07E056C76ED8DB2B69@bragi.housing.ufl.edu> References: <0E972CEE334BFE4291CD07E056C76ED8DB2B69@bragi.housing.ufl.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
In the last episode (Feb 20), Will Saxon said: > I am trying to use ntop in the 'border filter' mode to get some > statistics on that link, which is a gigE link typically seeing > 40-120Mbps sustained traffic (varies throughout the day). The > capturing interface is an intel pro/100+ server adapter, plugged into > another port on the switch set to mirror the gigE interface. > > Obviously there ought to be some drops due to serialization > differences, etc. But ntop is reporting upwards of 2/3 packets > dropped in the kernel and running tcpdump on that interface for any > length of time reports usually 50% drops in the kernel. You first need to determine what is being overloaded. Run top. Is ntop running at 100% cpu? If so, you'll need a faster machine. If it's close to 100%, bumping debug.bpf_bufsize might help. What are the user/system/irq CPU percentages while ntop is running? -- Dan Nelson dnelson@allantgroup.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030220170826.GO13096>