cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
439
Views
0
Helpful
8
Replies

When 2nd MPPP B-chan. connects, name resolution dies. OK again after discon

scott.rigby
Level 1
Level 1

I'm running a 2610 with IOS 12.1 using a Multilink PPP ISDN connection to the Internet as well as EasyIP and NAT.

When it's configured for 1 B channel, everything works fine.

When it's configured to bundle the two B-channels, as soon as you get the message "ISDN BRI 0/0:2 Connected to 1800123123" domain name resolution no longer works. If you enter the IP address of the site you're trying to get to though, it's fine.

If the 2nd B-channel goes down for some reason, name resolution instantly comes back.

I'm dialling into a cisco access router, so I would have expected it to work pretty smoothly. I'm a bit at a loss on this one - would appreciate some help.

Regards,

Scott... Melbourne, Australia

8 Replies 8

tepatel
Cisco Employee
Cisco Employee

Looks to me that router can't get to DNS server once the 2nd link is added to the bundle. So can you ping the dns server from the client?

Just verify that the PPP negotiation went fine when the 2nd link is added to bundle.

Thanks for the prompt reply.

You're right - I can't ping the DNS when the 2nd channel connects. I can ping another ISP's DNS, but trying to use their DNS address on my client didn't work.

I read somewhere else that the server needs to be running virtual templates. Can you confirm this?

Here's the PPP negotiation debug output - seems to be ok.

0:09:201863462912: %LINK-3-UPDOWN: Interface BRI0/0:2, changed state to up

00:09:204010946533: BR0/0:2 PPP: Treating connection as a callout

00:09:204028511588: BR0/0:2 PPP: Phase is ESTABLISHING, Active Open

00:09:201863463595: BR0/0:2 PPP: No remote authentication for call-out

00:09:201863462912: BR0/0:2 CHAP: Using alternate hostname xxxxxxxxxx

00:09:204035448316: BR0/0:2 LCP: O CONFREQ [Closed] id 5 len 26

00:09:206158430207: BR0/0:2 LCP: MagicNumber 0x308E5AAD (0x0506308E5AAD)

00:09:201863462912: BR0/0:2 LCP: MRRU 1524 (0x110405F4)

00:09:201863462912: BR0/0:2 LCP: EndpointDisc 1 Local (0x130C017065746572736F

6E73)

00:09:47: BR0/0:2 LCP: I CONFREQ [REQsent] id 1 len 28

00:09:47: BR0/0:2 LCP: AuthProto PAP (0x0304C023)

00:09:47: BR0/0:2 LCP: MagicNumber 0x8EB05770 (0x05068EB05770)

00:09:47: BR0/0:2 LCP: MRRU 1524 (0x110405F4)

00:09:47: BR0/0:2 LCP: EndpointDisc 1 Local (0x130A014D454C2D504F50)

00:09:47: BR0/0:2 LCP: O CONFACK [REQsent] id 1 len 28

00:09:47: BR0/0:2 LCP: AuthProto PAP (0x0304C023)

00:09:47: BR0/0:2 LCP: MagicNumber 0x8EB05770 (0x05068EB05770)

00:09:47: BR0/0:2 LCP: MRRU 1524 (0x110405F4)

00:09:47: BR0/0:2 LCP: EndpointDisc 1 Local (0x130A014D454C2D504F50)

00:09:47: BR0/0:2 LCP: I CONFACK [ACKsent] id 5 len 26

00:09:47: BR0/0:2 LCP: MagicNumber 0x308E5AAD (0x0506308E5AAD)

00:09:47: BR0/0:2 LCP: MRRU 1524 (0x110405F4)

00:09:47: BR0/0:2 LCP: EndpointDisc 1 Local (0x130C017065746572736F6E73)

00:09:47: BR0/0:2 LCP: State is Open

00:09:47: BR0/0:2 PPP: Phase is AUTHENTICATING, by the peer

00:09:47: BR0/0:2 PAP: O AUTH-REQ id 3 len 24 from "xxxxxxxxx"

00:09:48: BR0/0:2 PAP: I AUTH-ACK id 3 len 5

00:09:48: BR0/0:2 PPP: Phase is VIRTUALIZED

00:09:48: Vi1 MLP: Added link BR0/0:2 to bundle MEL-POP

00:09:49: %LINEPROTO-5-UPDOWN: Line protocol on Interface BRI0/0:2, changed stat

e to up

00:09:53: %ISDN-6-CONNECT: Interface BRI0/0:2 is now connected to 0198333320 MEL

-POP

Thanks - really appreciate your help.

P.S. the xxxxxxx's were just me protecting info.

Further info:

I'm using EasyIP to dial-in into the ISP and DHCP and NAT on the internal network. I just took a look at my router's config and noticed:

no ip name-server

Doubt that this is a problem because it works fine with a 1B channel connection.

Thanks....

Scott, whats the IOS you are running? Your symptoms sound like an old famous bug from 12.0.7T days (CSCdr45407 : cannot ping between routers when both channels are up) but you are only reporting DNS problem with both channels up, which is very unusual. So all IP traffic continues to work fine but its just the name resolution which stops working with both channels up ? And Yes you are right "no ip name-server" is not contributing in this problem.

~Zulfi

Thanks for your reply. I'm running 12.1(10) from memory. I thought about upgrading to the latest IOS but I don't have the right access here to download updates (and I bought the router second hand).

I think I might get a few other IP addrs (eg. their website) from the ISP and try pinging them. That will tell me whether it's only their DNS server that can't be pinged when 2 channels are up and not other machines within their network.

If I setup an ip name server (of another ISP) in the config, will that overide the DNS details that EasyIP installs upon connection? I noticed that I was able to ping it while the 2nd channel was up - but entering that DNS info into my client didn't allow me to resolve names. :(

Appreciate your help!

No one have any ideas here?

Would recommend opening a TAC case on this issue, access into the router would help troubleshooting it.

http://www.cisco.com/en/US/customer/support/index.html

Thanks, Mak.

What does your route table look like before and after the 2nd link comes up? What kind of routing are you doing? Config excerpt would help.