11-18-2012 05:39 AM - edited 03-07-2019 10:06 AM
Dear All,
This is my first post in this community, i hope you help me with this strange issue, i am facing a problem to tftp switch 2960s configuration to a tftp server, this problem does not happen with all switches, some of switches can tftp normally and others not, switches facing this issue are not specfic switches it is happen with different switches each time, they start to connecto to tftp after that the period (.....) start to appear after one ! , and show arp on these switches shows that they have an entry to the tftp server although switches are L2 switches and their managemnt VLAN are in totally different VLAN than tftp server , but the MAC is not the tftp machine MAC address, and this MAC is not one MAC it is two MACs changing with different switches, after searching on what these MACs belong to i found it belong to our core switches (NEXUS 7010) interfaces connecting each core to the other one and this link is for VPC keep alive
-here is our topology and IPs
- i searched over internet on release notes for 2960s switches with IOS (c2960s-universalk9-mz.150-1.SE.bin) version and found nothing
i hope someone can find an answer for this weared problem
thanks
11-18-2012 06:24 AM
Hi,
Yes, If TFTP server on another VLAN than switch there will be MAC of default gateway of Management VLAN. Because packet destined to other subnet any device sends packet to its GW.
Check connectivity between switch and server. Check trunks if those VLANs are allowed to pass through. And STP need to be checked.
Hope it will help.
Sent from Cisco Technical Support iPhone App
11-18-2012 07:11 AM
Dear Abzal
Thanks for your reply, but this is not the case , i know if the destination address is in different IP subnet than the source address then default gateway MAC address will be used as transition MAC address during commnication, my problem that i am not able to do tftp for switches, some do tftp and some don`t while i am taking backup from 21 switch, and show arp on these L2 switches shows arp record for tftp IP which is different vlan (although it should only show the arp for default gateway) , then i am not able to even ping tftp server , and to solve this i repeat claer arp cache till be able to ping tftp server again, and the MAC that appear in arp is not switch managment SVI MAC address or trunk physical MAC address, it is the MAC address of VPC keepalive link that appear in yellow color in the figure above.
Thanks again for your help
11-18-2012 07:43 AM
Hi,
Can you reach/ping your tftp from the 2960 switch?
Do you have the ip default gateway on the switch pointing to your core?
When you try to tftp does it give you this output !!!! Then ... Then !!! Then .....
Sent from Cisco Technical Support iPad App
11-18-2012 08:44 AM
Dear Khalid,
when tftp fail i can not and i find arp on the switch for 172.18.2.21 (tftp server) with mac of one of the keepalive links mac
yes i did point to the switch MGM default gateway on all switches which is CORE
it gives me !........
and error connecting tftp with ip 172.18.2.21
11-19-2012 03:40 AM
Dear All,
any idea what could be this issue related to is it NEXUS or 2960s switche or it is related to VPC or IOS bugs ?
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide