My original post:
Here is a snippet of /var/adm/messages:
Aug 18 10:39:04 ss04 unix: le0:
Aug 18 10:39:04 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 18 11:37:07 ss04 unix: le0:
Aug 18 11:37:07 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 18 11:48:05 ss04 unix: le0:
Aug 18 11:48:06 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 18 11:48:06 ss04 unix: le0:
Aug 18 11:48:06 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 18 12:04:02 ss04 unix: le0:
Aug 18 12:04:02 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 18 12:27:55 ss04 unix: le0:
Aug 18 12:27:55 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 20 10:46:00 ss04 unix: le0:
Aug 20 10:46:01 ss04 unix: No carrier - cable disconnected or hub link test disabled?
Aug 20 14:37:48 ss04 unix: le0:
Aug 20 14:37:48 ss04 unix: No carrier - cable disconnected or hub link test disabled?
No cable is disconnected and I am experiencing NO loss of connectivity
whatsoever. This has been going on for about three weeks now and once I
realized there was no actual loss of connectivity, I put looking into it
on my low priority list. I had a few moments today and could find nothing
out of the ordinary. I can't figure out what's generating this message. I
have searched SunSolve and found nothing useful.
During the time period when this started happening, there was no
software/hardware installed/deinstalled/reconfigured that I know of.
Any pointers to the possible source of the message would be appreciated so
I can quiet things down.
=====REPLIES=====
I heard from (in the order of when they appeared in my mailbox):
<bismark@alta.Jpl.Nasa.Gov>
<sweh@mpn.com>
<Michael.Neef@neuroinformatic.ruhr-uni-bochum.de>
<magi@csd.uwo.ca>
<gibian@stars1.hanscom.af.mil>
<medaglia@testmail.putnaminv.com>
<jharmon@telecnnct.com>
<Matthew.Stier@tddny.fujitsu.com>
<ctrapp@epidigm.mot.com>
<jcohen@paragren.com>
<chad@sequana.com>
<kamal@mcc.com>
<bobw@filmworks.com>
<vasu@center.oaep.go.th>
<anya@di.uoa.gr>
<Uwe.Dahm@pet.mpin-koeln.mpg.de>
Thanks to all who took the time to reply - there are truly some
knowlegeable folks on the list. Although I don't have the problem
completely solved, I believe I am heading in the right direction.
I first headed in the direction of the transceiver - a few replies
indicated that could be the cause of the problem. This machine (SS20) has
an 10baseT port in the back of the unit. Of course, the port could be
going bad, but read on! I replaced the cable patching me into the hub just
for kicks and that made no difference. I did notice, however, that the
data lights on the hub were pegged (bingo!) and occasionally, the
collision lights were flashing.
I think Kamal Kantawala <kamal@mcc.com> headed me in the right direction:
|Look at the collision count on your machine at that time:
|
|netstat -i 5
|
|It will poll every five seconds. I have noticed this problem at high
|collision instances. Also calculate your collision %, normally it should
|be low. I think 5% or less. It is a subjective number.
|
|One can count collision % by using the following command:
|
|netstat -n -i | grep ^le0 | awk '{print "Collision %: " $9 / $7 * 100}'
|
|Basically: Total Collision / Number of Output packets * 100.
|
|Kamal
When I was looking at this, I saw collision rates just over 10%. This is
extremely high. In fact, I think 10% is the magic number Sun uses to
indicate a problem. I got the 10% figure from the first line of output
from `netstat -i 5`. The first line gives you totals since boot time,
while the 5 second samples gives you immediate readings. The immediate
readings were closer to 22% (!!) Yipes!
I finally drew a correlation yesterday. This has been happening just prior
to our Novell 4.11 fileserver taking a dive. It is also preceeded by lots
(I mean lots) of NetBIOS traffic on the net. I'm sure this is not the end
of this story as I now have to go fight battles with those involved.
If any other significant events happen, I'll share them with the list.
Thanks for all your help.
-- John Cosimano Unix Systems Administrator The CNA Corporation Alexandria, VA USA cosimanj@cna.org
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:12:01 CDT