woody Posted May 16, 2007 Posted May 16, 2007 Sorry for the alarming heading, but something is going very wrong on my network. First symptoms were people not being able to print. This was weird becuase it was all the network prionters, yet there a several seperate print servers, and even the PCs that are directly connected over the network by an IP port stopped working. Sometimes I could ping these printers, sometimes intermittent ping returns, sometimes nothing. The HP printer in the offce was taking ages to start up, when I remove the network cable it gets to ready state prompto. Then there is the network switches. They all seem to be flashing weird, like every port is flashing simultaniously and this is on ALL of the switches around school. Those that have configuration pages I am unable to get to them, and I get 'Destination Unreachable' when I try and ping. The thing that throws me is that the switches must be working to some extent because I can log on, ping the network servers, get on the internet and send you this message, there have been no log on issues reported anywhere. So I am starting to think something sinister is happening, maybe something flooding the network on sertain ports? I am not a network engineer so I don;t know where to start investigating. Is there any software that can give me a picture of my network and where all the traffic is? Any help at this stage would be helpfull Thanks!
ThatBoringBloke Posted May 16, 2007 Posted May 16, 2007 Sounds like you have a loop. can you try isolating different segments of the network to pin it down?
Mr_T Posted May 16, 2007 Posted May 16, 2007 I agree with simongrahamuk have you pluged in any new network hardware or done any new recableing.
djm968 Posted May 16, 2007 Posted May 16, 2007 Agree, definately sounds like you have a loop somewhere.
marco84 Posted May 16, 2007 Posted May 16, 2007 Start by disconnecing one switch at a time from the network. When you have taken a switch down check the network traffic to see if things return to normal. Do this until you find the offending switch, then work through all ports patched into this switch.
woody Posted May 16, 2007 Author Posted May 16, 2007 Thanks Guys, will try isolating the problem. Geoff, what do you mean be STP?
Geoff Posted May 16, 2007 Posted May 16, 2007 Its a feature of managed switches. The full name is spanning tree protocol. Basically it detects loops and deactivates the looping port(s). If you search the forums you'll find a few threads on the subject, we've discussed it (and network loops) before.
woody Posted May 16, 2007 Author Posted May 16, 2007 Thanks Geoff, problem is at the moment I can't access the web configuration pages so can't switch it on
Geoff Posted May 16, 2007 Posted May 16, 2007 Most managed switches have serial and telnet based console access.
ConTheITGuy Posted May 16, 2007 Posted May 16, 2007 Then use the method as mentioned by Marco to track down and solve the problem, then switch on STP when it's all better Probably a kid playing silly beggars plugging one cable into two ports! We've had it here, as probably most others have. Most of the points aren't patched into the switch now if they're not being used. Andy
woody Posted May 16, 2007 Author Posted May 16, 2007 I am currently going round switching off the ones that I am able to at this time. I take it that when I turn off the offending switch, the others will immediately calm down and return to normal flashine of activity lights?
wesleyw Posted May 16, 2007 Posted May 16, 2007 Yep! Thats correct. However if you are have some throughput on the network then it's possible you have Rapid Spanning Tree/Spanning Tree Protocol working upto a point. But we have had broadcast loops like that to as others have already mentioned. Best to unplug all ports that aren't used to cut down on this ever happening again. Wes
woody Posted May 16, 2007 Author Posted May 16, 2007 Ok thanks guys, this would have to happen when one of my staff is off sick, and the other doesn't work wednesdays Will let you know how I get on
bossman Posted May 16, 2007 Posted May 16, 2007 I only enable STP on the particular port if i have for example extended that port by plugging in another switch. Other than that i have portfast enabled on all other ports which drops about 20-30 secs on packet transfers through the switch.
woody Posted May 16, 2007 Author Posted May 16, 2007 Right, I've found a cable plugged into two ports in a room where we recently took the computers out. Time to disconnect them I think! However, the switches are still flashing mad. Is there a method to bringing them back to normal, do I have to restart them all? I have tried restarting the one I think the offending cable was connected to, but that has had no effect.
Geoff Posted May 16, 2007 Posted May 16, 2007 If a switch has defaulted to hub mode due to the ARP storm you will need to reboot it to restore it to normal operation.
woody Posted May 16, 2007 Author Posted May 16, 2007 Well, sorted it out. There was another loop on a different switch in the Technology dept. Haven't found the actual loop yet as the ports aren't prperly labelled (I didn't do it) but I have disconnected that cable at the switch end and everything is back to normal. PHEW. Just a great big thanks to everyone who helped so quickly
Rozzer Posted May 16, 2007 Posted May 16, 2007 We had the exact same problem real pain tracing the loop back. We were thinking of putting on STP but turns out we cannot as we have channel bonding. Ross
DMcCoy Posted May 16, 2007 Posted May 16, 2007 I have no issues with RSTP and teamed links. RSTP is faster than STP so I don't get any issues with the time it takes to get DHCP addresses either. All procurves use it when you turn STP on afaik.
Guest kerrymoralee9280 Posted May 16, 2007 Posted May 16, 2007 Yeah, it happened to us 6 months ago. Network went nuts all of a sudden. Some kid had looped a network cable. We found the cable, and the culprit due to CCTV, but shortly after it happened again. NIGHTMARE. Anyhow, we enabled STP shortly after and not had the problem again.
philjones2000 Posted May 16, 2007 Posted May 16, 2007 Weirdly that happened to me exactly 1 week ago. Only difference was the amount of traffic shut down everything and I couldn't ping at all. Computer had been removed in ICT room, cable plugged in two sockets. Restart switches and everything went back. When I tried to recreate it though it wouldn't do it. Is there something special that needs to happen?
jumpinjamez Posted May 16, 2007 Posted May 16, 2007 Weirdly that happened to me exactly 1 week ago. Only difference was the amount of traffic shut down everything and I couldn't ping at all. Computer had been removed in ICT room, cable plugged in two sockets. Restart switches and everything went back. When I tried to recreate it though it wouldn't do it. Is there something special that needs to happen? We had this crossover problem for a week or so when the students found out it could slow everything down. We found is does take time for the network to come to a halt its not instant.
woody Posted May 17, 2007 Author Posted May 17, 2007 We had this crossover problem for a week or so when the students found out it could slow everything down. We found is does take time for the network to come to a halt its not instant. Yeah, for us it started with one printer not working properly, then two then all of them. After 24 hours everything was slow and printing was crippled.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now