cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
5362
Views
0
Helpful
25
Replies

[asr9k] interface up, but (l2transport) subinterface down

Carlos A. Silva
Level 3
Level 3

hi, all:

 

i was wondering if you could help out. I'm having a problem where i have an interface in up/up state, but when i configure an l2transport subinterface it doesn't come up/up. I was under the impression that as long as the parent interface is in up/up state, the subinterface will assume the same state.

 

 

RP/0/RSP0/CPU0:XXXXXX#show run int ten0/1/1/3               
Thu Jan 15 23:58:38.452 UTC
interface TenGigE0/1/1/3
 cdp
!

RP/0/RSP0/CPU0:XXXXXXX#show int ten0/1/1/3
Thu Jan 15 23:58:46.028 UTC
TenGigE0/1/1/3 is up, line protocol is up 
  Interface state transitions: 9
  Hardware is TenGigE, address is 5087.8933.5bdf (bia 5087.8933.5bdf)
  Layer 1 Transport Mode is LAN
  Internet address is Unknown
  MTU 1514 bytes, BW 10000000 Kbit 

[snip]

 

RP/0/RSP0/CPU0:XXXXXX#show run int ten0/1/1/3.2
Fri Jan 16 00:05:25.210 UTC
interface TenGigE0/1/1/3.2 l2transport
 encapsulation dot1q 2
!

RP/0/RSP0/CPU0:XXXXXXX#

 

TenGigE0/1/1/3.2 is up, line protocol is down 
  Interface state transitions: 1
  Hardware is VLAN sub-interface(s), address is 5087.8933.5bdf
  Layer 2 Transport Mode
  MTU 1518 bytes, BW 10000000 Kbit (Max: 10000000 Kbit)
     reliability Unknown, txload Unknown, rxload Unknown
  Encapsulation 802.1Q Virtual LAN,
    Outer Match: Dot1Q VLAN 2
    Ethertype Any, MAC Match src any, dest any
  loopback not set,

 

 

RP/0/RSP0/CPU0:XXXXXXX#show run int ten0/1/1/3.93
Thu Jan 15 23:59:17.707 UTC
interface TenGigE0/1/1/3.93 l2transport
 description PRUEBAS MARCATEL
 encapsulation dot1q 93
 rewrite ingress tag pop 1 symmetric
!

RP/0/RSP0/CPU0:XXXXXXX#

RP/0/RSP0/CPU0:XXXXXXXXX#show int ten0/1/1/3.93
Thu Jan 15 23:59:47.401 UTC
TenGigE0/1/1/3.93 is up, line protocol is down 
  Interface state transitions: 7
  Hardware is VLAN sub-interface(s), address is 5087.8933.5bdf
  Description: PRUEBAS MARCATEL
  Layer 2 Transport Mode
  MTU 1518 bytes, BW 10000000 Kbi

 

stuff like cdp works over the 10G link:

 

RP/0/RSP0/CPU0:XXXXXXXXX#show cdp ne
Fri Jan 16 00:02:32.009 UTC
Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge
                  S - Switch, H - Host, I - IGMP, r - Repeater

Device ID       Local Intrfce    Holdtme Capability Platform  Port ID
XXXXXXXXXX Te0/0/0/0        179     R          ASR9K Ser Te0/0/0/1       
ME38001        Te0/0/0/1        148     R S I      ME-3800X- Te0/1           
ME3600          Te0/1/1/3        150     S I        ME-3600X- Te0/1           
XXXXXXXXXXX Te0/1/0/0        120     R          ASR9K Ser Te0/0/0/1       
RP/0/RSP0/CPU0:XXXXXXXXX#

 

 

Seems weird to me, since we haven't had this happen before at all. Might be something very simple, but I cannot find it. On the other side I have a 3600 configured like this:

 

interface TenGigabitEthernet0/1

switchport trunk allowed vlan 2

switchport mode trunk

speed nonegotiate

 

 

Any ideas would be appreciated.

Thanks,

c.

 

25 Replies 25

Carlos A. Silva
Level 3
Level 3

Just in case:

 

RP/0/RSP0/CPU0:XXXXXXXX#show ver brief       
Fri Jan 16 00:10:40.743 UTC

Cisco IOS XR Software, Version 5.1.2[Default]
Copyright (c) 2014 by Cisco Systems, Inc.

ROM: System Bootstrap, Version 0.71(c) 1994-2012 by Cisco Systems,  Inc.

PE_CU uptime is 10 weeks, 5 days, 6 hours, 33 minutes
System image file is "disk0:asr9k-os-mbi-5.1.2/0x100305/mbiasr9k-rsp3.vm"

cisco ASR9K Series (Intel 686 F6M14S4) processor with 6291456K bytes of memory.
Intel 686 F6M14S4 processor at 2126MHz, Revision 2.174
ASR 9010 DC Chassis with PEM Version 2

4 Management Ethernet
88 GigabitEthernet/IEEE 802.3 interface(s)
16 TenGigE
16 DWDM controller(s)
16 WANPHY controller(s)
503k bytes of non-volatile configuration memory.
6111M bytes of hard disk.
12510192k bytes of disk0: (Sector size 512 bytes).
12510192k bytes of disk1: (Sector size 512 bytes).
RP/0/RSP0/CPU0:XXXXXXXX#

Is the subinterface "tied" with any other configuration element?
bridge domain with a BVI?
xconnect?
VPLS?
 

It was at some point, we ended up deleting the whole l2vpn section and starting from scratch.

 

In the end, a reload of the mod80 fixed the whole thing. not cool.

hi Carlos,

yeah that is not cool, and normally this doesn't happen, if the phy interface is up, the sub interfaces unless using BFD should be up as well, no question about it. What has likely happened is an interface creation error. There were some known ddts related to that in XR 51. but it is hard to say without traces what the precise issue is.

the show tech ethernet interfaces will probably help as it captures the traces from the creation/deletion and may have a clue there.

However after a reload that detail is gone, but if this happens again I would recommend capturing that trace file and open a tac case to relate it to either a known or new issue so it can be taken care of.

cheers

xander

actually, we have 2 mod80s on same box, doing the exact same thing. i don't think my co-worker reloaded both, so next step is seeing if the other is still doing the same thing and going straight to tac.

 

we've seen some funky stuff in the past 2 months on 5.1.2.

yeah hey carlos, if you can, look at XR513, it is rather solid and has a lot of fixes. It is also the EMR / extended maintenance release which longetivity and excellent smu/service pack support.

yeah please report back when if you have a repro or captures.

cheers!

xander

yeah, i've been thinking 5.1.3 for the past 2 months. we might have to look into that, after checking with tac about this last happening.

thanks for the advice.

while on this: for 5.1 and service packs, does it still have this behavior where every SP requires a reload and you can't mix and match SPs with SMUs (it's either or)?

 

Service packs still require a reload for installation that is correct.

You can apply individual smu's on top of a service pack also, I would recommend only doing that for time critical issues that need to be addressed "now", but for ease of use/management etc and all that it is best to stay with service packs when you decide to go down that path.

I am sure you are aware; CSM/cisco software manager is a great tool to help with the SMU scheduling and implementation too if you prefer to keep individuals as opposed to SP's.

 

cheers

xander

I ran into the same issue, below is my logs. I am using VIRL -

######################

RP/0/0/CPU0:NCS1#show run int gi0/0/0/1.10
Tue May  9 18:36:12.250 UTC
interface GigabitEthernet0/0/0/1.10 l2transport
 mtu 8918
!

#####################################

RP/0/0/CPU0:NCS1(config-if)#do sh ver brief
Tue May  9 18:31:08.561 UTC

Cisco IOS XR Software, Version 6.1.2[Default]
Copyright (c) 2016 by Cisco Systems, Inc.

#####################

RP/0/0/CPU0:NCS1(config-if)#do sh ip int bri
Tue May  9 18:30:31.753 UTC

Interface                      IP-Address      Status          Protocol Vrf-Name

GigabitEthernet0/0/0/1         unassigned      Up              Up       default
GigabitEthernet0/0/0/1.10      unassigned      Up              Down     default

#########################################

Initially the sub-interface was meant to be used under L2VPN xconnect config but even after removing it from L2vnp config; sub-interface is still down. Other side which is SR7750 however shows UP/UP.

########################################

Interface
-------------------------------------------------------------------------------
If Name          : Sub_NCS
Admin State      : Up                   Oper (v4/v6)      : Up/--

###################################

Thanks,

Amit

Hi Thuijs 

 

We have faced the same issue with L3 sub-interface and replaced the SPA card this has not resoleved the issue and we are running 5.1.0 

RP/0/RSP0/CPU0:KABANMALSAR02C9010#sh int des | i 0/5/1

Thu Oct  8 17:18:56.427 IST

[KTe0/5/1/0          down        down       

Te0/5/1/1          up          up          "TESTING"

Te0/5/1/1.100      up          down      

RP/0/RSP0/CPU0:XXXXXC9010#sh int des | i 0/5/1

Thu Oct  8 17:18:56.427 IST

Te0/5/1/0          down        down       

Te0/5/1/1          up          up          "TESTING"

Te0/5/1/1.100      up          down      

 

show tech ethernent interface is attched we have raised a tac case for the same ... 

let us know if you require anything 

hi kalsaria,

you have a sip700 with an ether spa? that is not a supported combination, you'd need to run ether on the tyhpoon or tomahawk linecards.

the sip700 can be used for pos, atm or channelized spa's only.

xander

We have tyhpoon card and on 1 SPA card it working and on second SPA card its not working. 

 

for module 0/5/0 -- sub-interface are in UP UP state 

for module 0/5/1 -- sub-interface are in UP Down state 

 

Hi Prakash,

are you using a MOD80-SE? have a look on DDTS: CSCuj28971

Thanks,

Ehsan

Yes we using MOD80-SE 

for the bug i have insufficient rights can you paste the content of the BUG

 

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: