I think we all agree that autonegotiation is evil, and should be avoided whenever possible. When you are looking for the root cause of the errors on your 3660, look at the speed and duplex settings for each device connecting to the etherstack hub. If one of those is miss-configured or possibly has a failing NIC, bad packets will be transmitted out all ports on the hub and will show up in the "show int f0/0" output on your router. Mike Braun -----Original Message----- From: Peter E. Fry [mailto:pfry@swbell.net] Sent: Tuesday, January 07, 2003 1:18 PM To: nanog@merit.edu Subject: Re: Weird networking issue. "David G. Andersen" wrote:
Rule number 1 with any ethernet: Check to make sure you have the duplex and rate statically configured, and configured identically on both ends of the connection. [...]
I'd like to thank Cisco for this piece of advice, as the only company incapable of manufacturing Ethernet equipment capable of autonegotiation. At least until 1999 or so. Yeah, there're a few others, all of which seemed to follow Cisco's lead. Nutty. Peter E. Fry "MMS <firstam.com>" made the following annotations on 01/07/03 14:22:30 ------------------------------------------------------------------------------ "THIS E-MAIL MESSAGE AND ANY FILES TRANSMITTED HEREWITH, ARE INTENDED SOLELY FOR THE USE OF THE INDIVIDUAL(S) ADDRESSED AND MAY CONTAIN CONFIDENTIAL, PROPRIETARY OR PRIVILEGED INFORMATION. IF YOU ARE NOT THE ADDRESSEE INDICATED IN THIS MESSAGE (OR RESPONSIBLE FOR DELIVERY OF THIS MESSAGE TO SUCH PERSON) YOU MAY NOT REVIEW, USE, DISCLOSE OR DISTRIBUTE THIS MESSAGE OR ANY FILES TRANSMITTED HEREWITH. IF YOU RECEIVE THIS MESSAGE IN ERROR, PLEASE CONTACT THE SENDER BY REPLY E-MAIL AND DELETE THIS MESSAGE AND ALL COPIES OF IT FROM YOUR SYSTEM." ==============================================================================
On Tue, 7 Jan 2003, Braun, Mike wrote:
I think we all agree that autonegotiation is evil, and should be avoided whenever possible. When you are looking for the root cause of the errors on
I don't agree. I have seen more problems generated by incompetence in trying to fix duplex/speed, than I have seen problems generated by autoneg not working properly. I am always amazed by the fact that very few people out there know that you have to lock duplex at BOTH ENDS of any given link for it to work properly. Generally, in a LAN environment with good quality switches and good network cards, autoneg works just fine. Yes, with 10/100 meg fiber/converters converters you should definately lock duplex, but in most other cases I recommend to leave the duplex setting to auto. Yes, cisco routers are notoriously bad at doing autoneg, but I blame that on cisco and not on autoneg. The el cheapo $50 desktop switches seem to hack autoneg just fine. -- Mikael Abrahamsson email: swmike@swm.pp.se
At 05:36 PM 1/7/2003, Mikael Abrahamsson wrote:
On Tue, 7 Jan 2003, Braun, Mike wrote:
I think we all agree that autonegotiation is evil, and should be avoided whenever possible. When you are looking for the root cause of the errors on
I don't agree. I have seen more problems generated by incompetence in trying to fix duplex/speed, than I have seen problems generated by autoneg not working properly.
I am always amazed by the fact that very few people out there know that you have to lock duplex at BOTH ENDS of any given link for it to work properly.
Generally, in a LAN environment with good quality switches and good network cards, autoneg works just fine. Yes, with 10/100 meg fiber/converters converters you should definately lock duplex, but in most other cases I recommend to leave the duplex setting to auto.
Yes, cisco routers are notoriously bad at doing autoneg, but I blame that on cisco and not on autoneg. The el cheapo $50 desktop switches seem to hack autoneg just fine.
Of all the gear I've worked with, from a wide variety of vendors, Cisco is the clear leader in gear that is incapable of successfully doing autonegotiation. I do hope they've improved this in newer products. The all time low point for them has to have been the 2924 switch. Putting a crossover cable between two 2924's yielded invariably BAD results. Now I can forgive engineers for not testing against every brand of router or host out there, but at LEAST test against another copy of the same box you're building. Not even something to blame on a QA engineer... this should have been tested before the box left the engineering benches. Connections to desktop computers and even servers are often better left set for autonegotiation. As people repatch connections to switches, it's easy to forget to reconfigure the switch. All that said, it's been my experience that when Cisco routers are involved, you really do have to force the interface settings or tempt fate.
I think we all agree that autonegotiation is evil, and should be avoided whenever possible. When you are looking for the root cause of the errors on
I don't agree. I have seen more problems generated by incompetence in trying to fix duplex/speed, than I have seen problems generated by autoneg not working properly.
I am always amazed by the fact that very few people out there know that you have to lock duplex at BOTH ENDS of any given link for it to work properly.
So thats human error not a problem with using forced settings, eliminate the human error and I think you'll see forced always works, autoneg sometimes works. (For future reference dont employ incompetent people to run your networks folks!)
Generally, in a LAN environment with good quality switches and good network cards, autoneg works just fine. Yes, with 10/100 meg fiber/converters converters you should definately lock duplex, but in most other cases I recommend to leave the duplex setting to auto.
Heh. I dont want to look at examples or find out what your experience is but in mine across a wide range of vendors its prone to problems.
Yes, cisco routers are notoriously bad at doing autoneg, but I blame that on cisco and not on autoneg. The el cheapo $50 desktop switches seem to hack autoneg just fine.
Have you looked at what autoneg is.. its horrible, a hack to help out the above incompetent engineers who dont know how to force duplex. .. well thats my opinion on the matter anyhow :) Steve
On Wed, 8 Jan 2003, Stephen J. Wilcox wrote:
So thats human error not a problem with using forced settings, eliminate the human error and I think you'll see forced always works, autoneg sometimes works. (For future reference dont employ incompetent people to run your networks folks!)
Problem with autoneg is that you always have to have manageble equipment and you always have to check both ends after changing anything. In an ISP environment that is generally not a problem luckily, apart from the equipment you connect to on the customer side, some customers insist on using cheapo stuff. Autoneg does add good things, especially on GigE. Autoneg on a GigE yields the most desireable effect of "link loss return", ie if you lose fiber link one way the link goes down at both ends.
Have you looked at what autoneg is.. its horrible, a hack to help out the above incompetent engineers who dont know how to force duplex.
Hmm, I might draw the same conclusion regarding automatic gear boxes on cars but I think I should not considering the situation in the US regarding that perticular issue :) Personally I think the idea with autoneg is really a good thing, why shouldnt two units advertise their capabilities and then act accordingly to what they both can do? We do the same on SMTP (EHLO) and so on. -- Mikael Abrahamsson email: swmike@swm.pp.se
participants (4)
-
Braun, Mike -
Daniel Senie -
Mikael Abrahamsson -
Stephen J. Wilcox