12-23-2013 10:46 AM - edited 03-04-2019 09:55 PM
Hi,
I was wondering if anyone has managed to sucessfully configure the EHWIC-4G-LTE-G network module for use on the O2-UK network?
I am succesfully registering to o2 (network 10) in the uk (country 234) without roaming (home):
#sh cellular 0/0/0 network
Current Service Status = Normal
Current Service = Packet switched
Current Roaming Status = Home
Network Selection Mode = Automatic
Mobile Country Code (MCC) = 234
Mobile Network Code (MNC) = 10
Packet switch domain(PS) state = Attached
Registration(EMM) state = Registered
And I have a really good 3G (UMTS) signal (-31dBm is like 5 bars):
#sh cellular 0/0/0 radio
Radio power mode = ON
Current RSSI(RSCP) = -31 dBm
LTE Technology Preference = AUTO
LTE Technology Selected = UMTS
But the profile won't stay active.
If I ping an address on the internet like 8.8.8.8, the default route and chat-script bring the modem online and if I very quickly view the profile it does become active:
#sh cellular 0/0/0 profile
Profile 1 = ACTIVE*
--------
PDP Type = IPv4
PDP address = 10.122.1.178
Access Point Name (APN) = mobile.o2.co.uk
Authentication = Unknown
Username: bypass
Password:
Primary DNS address = 82.132.254.2
Secondary DNS address = 82.132.254.3
However, after a few seconds it always reverts back to the inactive state:
#sh cellular 0/0/0 profile
Profile 1 = INACTIVE*
--------
PDP Type = IPv4
Access Point Name (APN) = mobile.o2.co.uk
Authentication = Unknown
Username: bypass
Password:
This is repeatable and I have tried vertigo, bypass and o2web as different APNs. Here is the profile configuration I'm using:
#cellular 0/0/0 lte profile create 1 mobile.o2.co.uk pap bypass password ipv4
I have also tried a variety of chat-scripts like:
chat-script modem "" "AT!CALL1" TIMEOUT 30
chat-script modem "" "ATDT*99*1#" TIMEOUT 60 "OK"
chat-script modem "" "ATDT*99*1#" TIMEOUT 60 "CONNECT"
chat-script modem "" "ATDT*99***1#" TIMEOUT 60 "CONNECT"
To no avail. Here is the debug output that does not look good:
*Dec 23 16:10:56.907: CHAT0/0/0: Attempting async line dialer script
*Dec 23 16:10:56.907: CHAT0/0/0: process started
*Dec 23 16:10:56.907: CHAT0/0/0: Asserting DTR
*Dec 23 16:10:56.907: TTY0/0/0: no timer type 1 to destroy
*Dec 23 16:10:56.907: TTY0/0/0: no timer type 0 to destroy
*Dec 23 16:10:56.907: TTY0/0/0: no timer type 2 to destroy
*Dec 23 16:10:57.063: TTY0/0/0: DSR was dropped
*Dec 23 16:10:57.063: tty0/0/0: Modem: READY->(unknown).
*Dec 23 16:10:58.063: TTY0/0/0: dropping DTR, hanging up
*Dec 23 16:10:58.063: TTY0/0/0: Async Int reset: Dropping DTR
For reference, here is the rest of the configuration:
interface Cellular0/0/0
ip address negotiated
encapsulation slip
dialer in-band
dialer idle-timeout 0
dialer string modem
dialer-group 1
async mode interactive
end
!
ip route 0.0.0.0 0.0.0.0 Cellular0/0/0
!
dialer-list 1 protocol ip permit
!
line 0/0/0
exec-timeout 0 0
script dialer modem
modem InOut
no exec
!
Any suggestions much appreciated.
Many thanks,
Jonny
12-31-2013 01:06 AM
For anyone who is interested, below is my working solution. The trick was to create a dialer interface with PPP encpasulation and join it to the cellular interface which has SLIP encapsulation:
!
ip dhcp pool LAN
network 192.168.1.0 255.255.255.0
default-router 192.168.1.1
dns-server 8.8.8.8 8.8.4.4
!
chat-script lte "" "AT!CALL1" TIMEOUT 30 "OK"
!
interface GigabitEthernet0/0
ip address 192.168.1.1 255.255.255.0
ip nat inside
ip virtual-reassembly in
duplex auto
speed auto
!
interface Cellular0/0/0
ip address negotiated
ip nat outside
ip virtual-reassembly in
encapsulation slip
dialer in-band
dialer pool-member 1
dialer-group 1
async mode interactive
routing dynamic
!
interface Dialer1
ip address negotiated
ip nat outside
ip virtual-reassembly in
encapsulation ppp
dialer pool 1
dialer idle-timeout 0
dialer string lte
dialer persistent
dialer-group 1
ppp authentication pap callin
ppp pap sent-username o2web password 0 password
ppp ipcp dns request
!
ip nat inside source list 1 interface Dialer1 overload
ip route 0.0.0.0 0.0.0.0 Dialer1
!
access-list 1 permit ip any
dialer-list 1 protocol ip permit
!
line 0/0/0
script dialer lte
modem InOut
no exec
transport input all
transport output all
rxspeed 100000000 <-- Max LTE download speed populated automatically
txspeed 50000000 <-- Max LTE upload speed populated automatically
!
As you can see the 10.x address assigned by o2 goes onto the Dialer interface instead:
GigabitEthernet0/0 192.168.1.1 YES NVRAM up up
Cellular0/0/0 unassigned YES NVRAM up up
Cellular0/0/1 unassigned YES unset down down
Cellular0/0/2 unassigned YES unset down down
Cellular0/0/3 unassigned YES unset down down
Dialer1 10.65.98.14 YES IPCP up up
NVI0 192.168.1.1 YES unset up up
These lines are also required but are the same as before:
#cellular 0/0/0 lte profile create 1 mobile.o2.co.uk pap o2web password ipv4
#cellular 0/0/1 lte profile create 1 mobile.o2.co.uk pap o2web password ipv4 <-- Necessary for some reason
#(config)interface cellular 0/0/0
#(config-if)#no shutdown
#(config)interface cellular 0/0/1
#(config-if)#shutdown
02-16-2015 03:55 AM
This configuration works better than the cellular based one, however, i'm working with the 881G-4G /LTE version, and when CEF is configured, packets from a PC connected to the LAN or routed port are erratic (bellow 50% success).
Turning off CEF, works fine.
I've tried several IOS including major and training version with the same result.
has anyone have this issue? help?
02-16-2015 12:03 PM
I have also experienced 50% packet loss over Cisco 4G modems before (normally after initial config) and never got to the bottom of it but a reboot always resolved it and it didn't normally return. Are you sure, re-enabling CEF brings the problem back? My packet loss was almost exactly 50% so I knew it was probably the router and not the radio link being flaky.
02-18-2015 05:31 AM
Hi, Jonathan.
Thank you very much for your input.
By rebooting the router it doesn't improve. The symptom is the same.
If CEF is enabled, the PING responds close to 50% (one ok, one nok).
CEF improves routing performance and with 4G bandwitdh available, when it's disabled (CEF), it could mean some performance degradation.
I dind't open a case yet, but I'm affraid that Cisco will recomend to remove dialer configuration.
Dialer configuration allows "allways on" configuration in 4G access. For a provider, that's relevant.
Regards,
03-18-2015 03:54 AM
Hi Santosnun,
Did you get anywhere with the case? I have the same thing and reboot does not help. When I turn on debug ip cef drop I see CEF is dropping the packets because of missing adjacency.
Many thanks in advance,
Bud
06-22-2015 06:16 AM
Yes in the end this was resolved by going to IOS 15.3 on the router; it appears to be a bug on IOS 15.2 which effects CEF.
06-22-2015 07:42 AM
We're using 15.4 and the problem persists...
02-11-2019 11:08 PM
For all those struggling with 50% packet loss, no routing, no ICMP through an LTE cellular interface, on a 2901, 1901, 1921, etc, etc, disable IP CEF on the cellular interface. It is both a global and interface specific command. It may not be a bug, it may be that your carrier is pushing CEF packets and thus mucking up your routing.
Probably affects all IOS(since it may be a feature !bug)
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