Here's my original posting:
>
> Several times a day the following message appears on the console:
> rpc.etherd: bad lnth 42 dst ffffffffffffsrc 08002008ef1d
>
> The number 08002008ef1d is the ethernet address of the host.
> The message isn't logged by syslog.
> Also, rpc.etherd is using 1-2% of the cpu cycles all the time.
> I don't remember this behavior before.
>
> Any idea what's happening?
I received replies from 8 people, and the concensus is that there
is no real problem. The most concise answer came from
chris@rufus.state.COM.AU:
>It's nothing much to worry about. rpc.etherd sits at a "funny" layer
just
>above the ethernet/loopback drivers. What you are seeing are loopback
>packets from the host that by ethernet standards are indeed "runts".
>However, these aren't being pushed out onto the ethernet. In fact 42
>looks like the length of an ARP packet. It's probably resolved
internally
>and thus never sent out onto the network.
>rpc.etherd chews CPU proportionally to the amount of traffic on the
network.
There was an inquiry about this by Eckhard R"uggeberg
(eckhard@ts.go.dlr.de)
some time ago, and he kindly sent me his summary. It included this
fragment:
>We reported the bug to Sun many
> months ago, and they just this week sent us a patch to try out. I
don't
> think it has an official patch number yet.
(shj@ultra.com (Steve Jay {Ultra Unix SW Mgr}) )
I have set my sources to work to see if we can find the official bug
and patch number for this.
Thanks to:
<aahvdl@eye.psych.umn.edu>
<tom@yac.llnl.gov>
<andataco!andataco.com!louis@UCSD.EDU>
<dla@se05.wg2.waii.com>
<miker@il.us.swissbank.com>
<eckhard@ts.go.dlr.de>
<matis.ingr.com!amir@matis.uucp>
<chris@rufus.state.COM.AU>
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:23 CDT