Loop errdisable
Rearden
Member Posts: 222
in CCNA & CCENT
Hey. Wow, it's been a while since I've been here. I passed my CCNA back in August, but I've been so busy with school that I haven't had time to start studying CCNP material much. Ah well, I'm only a year and two thirds of semester away from graduation (Computer Science).
How has everyone been? Did mike try the CCIE lab again? I wish him the best of luck either way.
Anyway what made me stop in was an issue that cam up at work yesterday and I'm not really sure why this happened. Maybe someone here can explain it. Here's the Reader's Digest version:
A student calls up saying his port isn't working. Helpdesk sends it down to networking because his machine doesn't appear to work in his roommates port either, and now his roommate's port doesn't work at all. The caller said that the trouble started right after he booted wit an Ubuntu liveCD. I go down there and his setup is network port -> hub (that's actual layer 1 hub) -> pc, xbox
So, I figure the hub somehow pissed off the switch (it's a Catalyst 3560G PoE) and the switch turned it off. I go log into the switch and sure enough it was errdisabled. Now, since we don't have any other way to check why it was errdisabled, I went and checked the tacacs logs (tacas is really really cool, by the way) and they say it was. Here are the actual entries, IP addresses obfuscated because I'm paranoid even though they're 10.x.x.x addresses
That seemed rather unusual since he claims he never had the hub connected to both ports at the same time. Also, the time stamps are different so it doesn't seem like he had the thing plugged into two ports at the same time. I presumed it was the hub that was causing all of this trouble, reset his port and told him to go buy a desktop switch.
However, while testing, the second I plugged that hub in, the port shut right down again. I just don't see how the switch saw the keepalive message, which Cisco says causes loop disables out the same port it sensed, unless the hub is somehow broken and sending messages out every port including the one it cam in on.
But, I still don't understand how his port ended up in a loop error state or why it would make a difference what OS he used. I'm a bit skeptical that the problem really started right when he booted that liveCD. Anyone have any ideas on this one? It just seems odd.
How has everyone been? Did mike try the CCIE lab again? I wish him the best of luck either way.
Anyway what made me stop in was an issue that cam up at work yesterday and I'm not really sure why this happened. Maybe someone here can explain it. Here's the Reader's Digest version:
A student calls up saying his port isn't working. Helpdesk sends it down to networking because his machine doesn't appear to work in his roommates port either, and now his roommate's port doesn't work at all. The caller said that the trouble started right after he booted wit an Ubuntu liveCD. I go down there and his setup is network port -> hub (that's actual layer 1 hub) -> pc, xbox
So, I figure the hub somehow pissed off the switch (it's a Catalyst 3560G PoE) and the switch turned it off. I go log into the switch and sure enough it was errdisabled. Now, since we don't have any other way to check why it was errdisabled, I went and checked the tacacs logs (tacas is really really cool, by the way) and they say it was. Here are the actual entries, IP addresses obfuscated because I'm paranoid even though they're 10.x.x.x addresses
Feb 14 09:35:51 <IP> 1676: 4w0d: %PM-4-ERR_DISABLE: loopback error detected on Gi0/10, putting Gi0/10 in err-disable state Feb 15 14:30:05 <IP> 1748: 4w1d: %PM-4-ERR_DISABLE: loopback error detected on Gi0/4, putting Gi0/4 in err-disable state
That seemed rather unusual since he claims he never had the hub connected to both ports at the same time. Also, the time stamps are different so it doesn't seem like he had the thing plugged into two ports at the same time. I presumed it was the hub that was causing all of this trouble, reset his port and told him to go buy a desktop switch.
However, while testing, the second I plugged that hub in, the port shut right down again. I just don't see how the switch saw the keepalive message, which Cisco says causes loop disables out the same port it sensed, unless the hub is somehow broken and sending messages out every port including the one it cam in on.
But, I still don't understand how his port ended up in a loop error state or why it would make a difference what OS he used. I'm a bit skeptical that the problem really started right when he booted that liveCD. Anyone have any ideas on this one? It just seems odd.
More systems have been wiped out by admins than any cracker could do in a lifetime.
Comments
-
EdTheLad
Member Posts: 2,111 ■■■■□□□□□□
Either you have a faulty cable,faulty hub port or faulty hub.The tx and rx being shorted somewhere.
You could try disabling the keepalive on the switchport as a workaround.Networking, sometimes i love it, mostly i hate it.Its all about the $$$$ -
Rearden
Member Posts: 222
My guess is that the guy had a bad hub. I tried it with a known good patch cable and still got the same errors. Thanks for the confirmation
More systems have been wiped out by admins than any cracker could do in a lifetime. -
bighornsheep
Member Posts: 1,506
Those are common if you have a speed/duplex mismatch. The hub will operate at half-duplex whereas the 3560 will try and negotiate full-duplex, I am not sure if it depends on different switch IOS, but I have seen on other models that the auto/auto will "seem" to have negotiated 100/Half, but eventually you seen a bunch of CRC errors and frames are dropping, very soon you start getting those errdisable states on the switch side.
I don't think hubs can auto-negotiate, try forcing the port the hub is connected on to be 100/Half (or 10/Half if it's a 10Mb hub), turn off keepalive (as Ed suggested) and see what happens.Jack of all trades, master of none