Sewer-rat
11-03-2010, 11:55 PM
I have come across this issue twice at my job this month and found it interesting enough to post in the switching threads. I had a call from a client who had a 5 circuits bounce for a short period and come back up. When I started tracing the vlans path I found that every trunk link the vlans took bounced. The logs showed the interface going down and coming back up all with the same time stamp across the switches. All the sites were in the same city but the vlans were routed through 4 different cities back to a main site and finally I found it...Between two Distribution switches connecting two cities there was an error:
%DTP-5-DOMAINMISMATCH: Unable to perform trunk negotiation on port Gix/x because of VTP domain mismatch.
In my environment we don't use VTP so to speak since when we design circuits for clients everything is added manually and our switches run in Transparent mode. I looked into it and sure enough the trunks were hard coded for trunking but negotiation was left on along with each switch being in a different VTP domain. I sent it up for a local engineer to review but he noted that we do not use VTP and that the trunks were hard coded. Ok...so why did this happen?? :confused:
A tiered engineer stepped in and actually opened a TAC case with Cisco to verify what happened here. Cisco noted that even if you have switches running in transparent mode with hard coded trunks mismatched domains will still cause DTP negotiation failures if the negotiation is left on. Sure enough just yesterday I had the same issue, same client but a different site! Needless to say we didn't need a TAC case this time :rolleyes: Figure I would share this with the class :D
%DTP-5-DOMAINMISMATCH: Unable to perform trunk negotiation on port Gix/x because of VTP domain mismatch.
In my environment we don't use VTP so to speak since when we design circuits for clients everything is added manually and our switches run in Transparent mode. I looked into it and sure enough the trunks were hard coded for trunking but negotiation was left on along with each switch being in a different VTP domain. I sent it up for a local engineer to review but he noted that we do not use VTP and that the trunks were hard coded. Ok...so why did this happen?? :confused:
A tiered engineer stepped in and actually opened a TAC case with Cisco to verify what happened here. Cisco noted that even if you have switches running in transparent mode with hard coded trunks mismatched domains will still cause DTP negotiation failures if the negotiation is left on. Sure enough just yesterday I had the same issue, same client but a different site! Needless to say we didn't need a TAC case this time :rolleyes: Figure I would share this with the class :D