home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.lans.ethernet
- Path: sparky!uunet!grebyn!daily!rich
- From: rich@grebyn.com (Richard Lawrence)
- Subject: Re: le0: Transmit late collision - net problem?
- Message-ID: <1992Dec22.035502.12965@grebyn.com>
- Keywords: SunOS, twisted-pair Ether, Synoptics
- Organization: Grebyn Timesharing
- References: <1992Dec16.181735@is.morgan.com>
- Date: Tue, 22 Dec 1992 03:55:02 GMT
- Lines: 30
-
- In article <1992Dec16.181735@is.morgan.com> sergei@is.morgan.com (Sergei Poliakoff) writes:
- >
- >We are tracking an uncommon condition : having enabled a ledebug
- >in SunOS kernel, we discovered fair amount of
- >
- >le0: Transmit late collision - net problem?
- >
- >appearing in syslogs of stations on just one subnet. I can reproduce it on
- >any workstation on that segment by forcing massive transfers from the
- >workstation, which points me towards some network-related rather then
- >system related explanation. It never occurs on other subnets.
-
- Which type of UTP cabling are you using? You may have stretched past
- your delay budget...also, note that even if you force if from every
- station and get late collisions, it STILL might be another station that
- is attempting to transmit erroneously (in other words, you jam the
- wire with traffic from station A, and sooner or later station B wants to
- send a packet for some reason (ARP, whatever) and collides).
-
- If you have managment options on your Synoptics concentrators (you
- didn't specify which ones you have) you can get diagnostic information
- from the concentrator itself. It will tell you to the port level about
- conditions such as late collisions.
-
- >
- --
- Rich Lawrence, Synoptics Systems Engineer DOD#9630 rich@grebyn.com
- '92 Seca II "Yuri" CI$:71101,2272 GEnie:R.LAWRENCE14
- 216 years ago my government gave me certain inalienable rights.
- Since then, they've been trying to correct their mistake.
-