How can that be possible? I have an all switched network (24 Pure data + 2 small netgear switches) and from time to time I would see the colisions light blink in the netgear. A show interface command on the router would show colisions on that interface too. That just make no sense.
Colisions in an all switched network.
- Thread starter Dark
- Start date
Collisions are normal, and required in a Half-Duplex Network, whether it is switched or not does not matter, however they will be reduced signifigantly as only the Switch sending data can colide with the workstation Sending data.
You will not see collisions in a Full Duplex network (unless you have bigger issues)
You will not see collisions in a Full Duplex network (unless you have bigger issues)
Xanthars got it.
collisions are a normal operation of half-duplex ethernet (switched or not). If you see collisions on full-duplex ports you have something mis configured or a cabling problem or a flaky NIC/drivers. In fact you should NEVER see any kind of errors on a full-duplex link.
collisions are a normal operation of half-duplex ethernet (switched or not). If you see collisions on full-duplex ports you have something mis configured or a cabling problem or a flaky NIC/drivers. In fact you should NEVER see any kind of errors on a full-duplex link.
I just finally tracked down my collisions with my 3com router (10bT hub buildt in). The collision light would just constantly blink once a second. Turns out that the 3com is picking up false collisions. Who would have guessed it.
also...check the configuration of the fastE port on the router. If it is cisco it defaults to auto, but for a router or other important resource you should force the speed and duplex on both sides of the link.
One more tip...ALWAYS set both sides of the link to the same operation. If you want to use autonegotiation the set both sides to auto. Or force both sides. DONT set one side of the link to 100/full and leave the other side at auto. This can cause duplex mismatches.
One more tip...ALWAYS set both sides of the link to the same operation. If you want to use autonegotiation the set both sides to auto. Or force both sides. DONT set one side of the link to 100/full and leave the other side at auto. This can cause duplex mismatches.
Thx for all your reply all. I'm not using Halfduplex at all. All the nic are set to auto and the lights on the switch show that FDX is used. Regarding the router (cisco), I could not set it as fullduplex, I don't know why but it keeps saying command unknown!!!! (config t, int e0, media-type 10baseT, full-duplex) the last command is always refused. I'm using IOS 12.0 and up to 11.x that was the command i've been using to set the half-full.
Ahh, 1600 only has a 10/half port. collisions are perfectly normal on that port. Again, you shouldn't have any collisions on the full-duplex ports.
Dumb me. I never read that on the product documentation. Is it by experience that you guys saying that or I missed it somewhere?
Experience with the 1605 in particular (but it holds true for the 160X series as well). I used to work in HiCap (tech support for dedicated customers from modems to DS-3s) and all of our T-1 customers received a 1605-R as part of the deal. I've probaby set up about 60 of these.
Heh, here's a funny one and definitely off this topic, but what the heck.
Today we turned up another data T-1 to a remote POP. I thought it was coming in from Ameritech but it turned out we got it from Time Warner instead (have I ever told anyone how much I LOVE TW TelCo?!?!?!
). Anyways...so I had configured this channel group on one interface on this 7206 VXR (NPE 300 btw) to reserve it so that no one else would take it because it was the next available timeslot on that DS-3.
No problem right? Wrong. This router does NOT like me. I swear it has something personal against me. So, I conf t, int se2/1/11:0, no ip address, end, write. All should be good. Then I go and configure the channel group up on the other DS-3 card channel (same router now folks), then configure that serial interface and add the load sharing crap and add the network interface to the ospf area so that we can load balance on the two T-1s.
Doesn't work. In fact, connectivity to to the remote router gets to spotty at times that TACACS+ won't allow me to issue commands. FSCK!
After about 30 minutes of farting around I finally issue, which I should've done right off (duh!) sh ip route ospf <Process ID>. And lo and behold it has two interfaces with the same IP number. WTF!! Easily fixed, but this has to be like the seventh or eighth time this router has not added or deleted something even though I know I issued the command (trust me I watch it like a hawk now).
Just thought I'd share one with 'ya.
Today we turned up another data T-1 to a remote POP. I thought it was coming in from Ameritech but it turned out we got it from Time Warner instead (have I ever told anyone how much I LOVE TW TelCo?!?!?!
). Anyways...so I had configured this channel group on one interface on this 7206 VXR (NPE 300 btw) to reserve it so that no one else would take it because it was the next available timeslot on that DS-3.No problem right? Wrong. This router does NOT like me. I swear it has something personal against me. So, I conf t, int se2/1/11:0, no ip address, end, write. All should be good. Then I go and configure the channel group up on the other DS-3 card channel (same router now folks), then configure that serial interface and add the load sharing crap and add the network interface to the ospf area so that we can load balance on the two T-1s.
Doesn't work. In fact, connectivity to to the remote router gets to spotty at times that TACACS+ won't allow me to issue commands. FSCK!
After about 30 minutes of farting around I finally issue, which I should've done right off (duh!) sh ip route ospf <Process ID>. And lo and behold it has two interfaces with the same IP number. WTF!! Easily fixed, but this has to be like the seventh or eighth time this router has not added or deleted something even though I know I issued the command (trust me I watch it like a hawk now).
Just thought I'd share one with 'ya.
Damaged,
what no NPE400?
On the OSPF note, you might want to let the LSA age out of the LDdabase. I think default is 30 min update and 60 min age. Or you could restart the ospf process (carefull here)
Call tac, let them deal with it.
what no NPE400?
On the OSPF note, you might want to let the LSA age out of the LDdabase. I think default is 30 min update and 60 min age. Or you could restart the ospf process (carefull here)
Call tac, let them deal with it.
Heh, no. Sorry, no NPE 400. Like we need it. This router currently has 2 PA-2MT3+ cards, one single MT3, and one 8 port channelized T-1 card, with a total of 118 interfaces. Wee! Nice when you do a sh run. Max CPU usage approaches something like 8%. These are mostly T-1 customers, statically routed, but we do have some POPs on there running OSPF.
Thanks goodness for ip unicast verify reverse-path <access list number> or I'd have individual access lists for each customer and the config would be truly unruly to view. This way I can use two generic access lists and apply to all T-1 customer interfaces <whew>.
Spidey, I would, but the danged thing only seems to do it for me. No one else seems to have a problem with it. Granted this is kinda my baby and I definitely get more face time on it, but...grrrrr. We do have a service contract on it, so I think I'll start keeping some notes on that and fire off and email to them and see what they can see.
Need more caffeine...NOW! ;P
Thanks goodness for ip unicast verify reverse-path <access list number> or I'd have individual access lists for each customer and the config would be truly unruly to view. This way I can use two generic access lists and apply to all T-1 customer interfaces <whew>.
Spidey, I would, but the danged thing only seems to do it for me. No one else seems to have a problem with it. Granted this is kinda my baby and I definitely get more face time on it, but...grrrrr.
We do have a service contract on it, so I think I'll start keeping some notes on that and fire off and email to them and see what they can see.Need more caffeine...NOW! ;P
on the show run note. Yeah, the configs get pretty big.
Learn to use regular expressions for filtering: show run | include serial
or sh run serial 2/0:20
Learn to use regular expressions for filtering: show run | include serial
or sh run serial 2/0:20
Oh, I do use those, but sometimes I just need to scroll through 'cause I can't remember 'zackly what to "grep" for there.
Now, If we could just get Redback on the ball with that.
Now, If we could just get Redback on the ball with that.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 411
-
Facebook
Twitter