07-13-2015 01:15 AM
The 887VAG+R7-K9 router is giving me much grief.
This time it's the unstable ezVPN tunnel. Sometimes there seems to be an endless loop where a successful tunnel build-up seems impossible. A power-cycle might help, or a erase config->replace config does the trick.. The behavior is very erratic at best!
I am using several 887VAG+R7 units with different mobile operators in our country. I am hesitant to put more in the field because of this particular issue. All the units connect to some sort of ASA router (either a 5505 or a 5512-X). IOS image versions differ and do not seem to make any difference on the behavior.
This is what you see at the ASA when things go south:
This keep looping over and over... But the connection DID work fine at first. This just randomly manifested itself! It's driving me nuts! Our vast amount of 887VA units (ADSL) or 881 units (ethernet) never exhibit this behavior on the same ASA's.
This is the configuration of the 887VAG+R7 in question:
version 15.2 no service pad service tcp-keepalives-in service tcp-keepalives-out service timestamps debug datetime msec service timestamps log datetime msec service password-encryption ! hostname User160 ! boot-start-marker boot-end-marker ! no logging buffered no logging console ! aaa new-model ! aaa session-id common memory-size iomem 10 clock timezone Paris 1 0 clock summer-time DST recurring last Sun Mar 2:00 last Sun Oct 3:00 ! no ip source-route ! no ip bootp server ip domain lookup source-interface Cellular0 ip domain name vodafone.nl ip name-server 8.8.4.4 ip cef no ipv6 cef ! multilink bundle-name authenticated chat-script hspa-R7 "" "AT!SCACT=1,1" TIMEOUT 60 "OK" license udi pid C887VAG+7-K9 sn FCZ1906C2S2 ! controller VDSL 0 shutdown ! controller Cellular 0 gsm gps mode standalone gsm gps nmea ! ip ssh version 2 ! crypto isakmp keepalive 60 5 periodic ! crypto ipsec client ezvpn HOI connect auto group X_Tunnel_Group key <omitted> mode network-extension peer 1.2.3.4 username User160 password User160 xauth userid mode local ! interface Ethernet0 no ip address shutdown ! interface ATM0 no ip address shutdown no atm ilmi-keepalive ! interface FastEthernet0 no ip address ! interface FastEthernet1 no ip address ! interface FastEthernet2 no ip address ! interface FastEthernet3 no ip address ! interface Cellular0 ip address negotiated encapsulation slip dialer in-band dialer idle-timeout 0 dialer string hspa-R7 dialer-group 1 async mode interactive crypto ipsec client ezvpn VRI ! interface Vlan1 ip address 172.16.83.206 255.255.255.248 no autostate crypto ipsec client ezvpn HOI inside ! ip forward-protocol nd no ip http server ip http access-class 10 ip http authentication local ip http secure-server ! ip route 0.0.0.0 0.0.0.0 Cellular0 ! ip sla 1 icmp-echo 10.40.1.161 source-interface Vlan1 frequency 300 ip sla schedule 1 life forever start-time now access-list 10 remark CCP_ACL Category=17 access-list 10 remark VPN subnet access-list 10 permit 172.18.0.0 0.0.0.255 dialer-list 1 protocol ip permit no cdp run ! control-plane ! line con 0 no modem enable line aux 0 line 3 exec-timeout 0 0 script dialer hspa-R7 modem InOut no exec rxspeed 21600000 txspeed 5760000 line 6 modem InOut no exec transport input all transport output all stopbits 1 speed 4800 line vty 0 4 access-class 10 in privilege level 15 transport input ssh ! end
These things really exhibit weird and random behavior at times which I can only blame the router for. Other variables (like mobile sims or operator brand) make no difference whatsoever. Of course the configuration could be at fault, however, the fact it leads to inconsistent behavior is something that strikes me as very odd. The only thing that keeps me clinging to these units is the fact that they support ezVPN with Network Extension Mode.
Has anyone seen this behavior and have they found a solution for it?
If you could advise it would be greatly appreciated.
Thanks in advance!
07-14-2015 09:10 PM
Hi,
From the messages it looks like the VPN peers are not in sync, the Router is sending encrypted packets fro the ASA but it has no ipsec SA for the same.
Could you check if the isakmp keep-alives are applied on both sides and if you can simultaneous debugs on both sides to see why the tunnel is going down.
HTH
Abaji.
07-15-2015 06:14 AM
Hallo Abaji,
Thank you for your insight and reply!
The DPD is enabled on both ends of the tunnel. The ASA has the DPD enabled in the tunnelgroup and the 887VAG+R7 has the line "crypto isakmp keepalive 60 5 periodic". I do believe that's the correct way to enable DPD... correct me if I am mistaken.
I did notice the confidence interval is different between the ASA and the 887VAG+R7. One is 30 seconds (ASA) and the 887VAG+R7 has 60 seconds. Could this cause this kind of trouble? Perhaps I am making the wrong assumption that one side concludes after 30 seconds a tunnel has dropped and the other side concludes it after 60 seconds. That shouldn't be a problem, right?
If it was a DPD issue however, I would suspect there being a lingering SA on the ASA. When looking at the ASA active SA's I noticed one IKE peer (the failing 887VAG+R7) with the state: "AM_TM_INIT_MODECFG_V6H". The search to this particular state lead me to another forumpost which suggests the problem might be fragment/MTU related. Which could be it.. I am not sure.
What I do know is that the 887VAG+R7 thinks Phase 1 is completed while the ASA doesn't. That could very well be because of fragmented packets.
07-20-2015 08:34 AM
You are correct. 887VAG+R7 thinks Phase 1 is completed while the ASA doesn't.
Need to check debugs on both ends.
HTH
Abaji.
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