Unsolved

This post is more than 5 years old

27 Posts

98612

August 11th, 2014 16:00

MLAG to VLT mesh connection

Hi,

We have a pair of S4810 and N4032 switches in our lab. So we've configured vlt and mlag with a mesh topology. 

Although it seems working, following lacp ungroup/group messages appear on the force10 console. As soon as I disabled/shutdown one of the lacp port then it stops dropping any new message.

4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-UNGROUPED: PortChannel-011-Ungrouped: Interface Te 0/1 exited port-channel 11.

4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-GROUPED: PortChannel-011-Grouped: Interface Te 0/1 joined port-channel 11.
4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-UNGROUPED: PortChannel-011-Ungrouped: Interface Te 0/1 exited port-channel 11.
4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-GROUPED: PortChannel-011-Grouped: Interface Te 0/1 joined port-channel 11.
4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-UNGROUPED: PortChannel-011-Ungrouped: Interface Te 0/1 exited port-channel 11.
4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-GROUPED: PortChannel-011-Grouped: Interface Te 0/1 joined port-channel 11.
4d17h13m: %STKUNIT0-M:CP %LACP-5-PORT-UNGROUPED: PortChannel-011-Ungrouped: Interface Te 0/1 exited port-channel 11.

Any help would be appreciated. Also, partial config on both n4032 and s4810 config can be seen below.

N4032

interface Te1/0/1

channel-group 20 mode active

description "MLAG-to-VLT"

exit

!

interface Te1/0/2

channel-group 20 mode active

description "MLAG-to-VLT"

exit

interface port-channel 20

description "N4032 to S4810"

spanning-tree disable

switchport mode trunk

vpc 20

exit

feature vpc

vpc domain 1

peer-keepalive enable

 

FORCE10

lacp ungroup member-independent vlt

!

protocol spanning-tree rstp

 no disable

 hello-time 1

 max-age 6

 forward-delay 4

!

vlt domain 1

 peer-link port-channel 127

 back-up destination 10.80.30.15

 primary-priority 8192

 system-mac mac-address 00:01:e8:8b:e9:7e

 unit-id 1

 peer-routing

interface TenGigabitEthernet 0/0

 no ip address

!

 port-channel-protocol LACP

  port-channel 11 mode active

 no shutdown

!

interface TenGigabitEthernet 0/1

 no ip address

!

 port-channel-protocol LACP

  port-channel 11 mode active

 no shutdown

interface Port-channel 11

 no ip address

 switchport

 vlt-peer-lag port-channel 11

 no shutdown

!

interface Port-channel 127

 description S4810_OSPF-Aggregation-1_0/48_S4810_OSPF-Aggregation-1_0/52

 no ip address

 channel-member fortyGigE 0/48,52

 no shutdown

27 Posts

August 12th, 2014 12:00

I've changed the ports on both force10 switches so now they are configured on ports 10&11 instead of 0&1

Although all the leds are green and interfaces are up I've replaced one of the cable just in case as you suggested. However messages still appear on the console.(on both force10 console)

N4032-top#show vpc brief

.....(removed the other details)

VPC id# 20

-----------

Interface...................................... Po20

Configured VLANs............................... 1,30,104,1001

VPC interface state............................ Active

Local Members        Status

-----------------    ------

Te1/0/1              Up

Te1/0/2              Up

Peer Members         Status

----------------     ------

Te1/0/1              Up

Te1/0/2              Up

s4810-top#sh interfaces port-channel 11 brief

Codes: L - LACP Port-channel

      O - OpenFlow Controller Port-channel

   LAG  Mode  Status       Uptime      Ports

L   11   L2L3  up           00:11:34    Te 0/10    (Up)

                                       Te 0/11    (Up)

Similar outputs are shown from the N4032 and S4810 pairs.

27 Posts

August 12th, 2014 12:00

FTOS version is 9.5

Since I've moved the cables to ports 10&11 for portchannel 11, messages are no longer being reported as te 0/1.

Current connection between N4032 mlag and S4810 VLT

S4810-top port 11 N4032-top port 1

S4810-top port 10 N4032-btm port 2

S4810-btm port 11 N4032-btm port 1

S4810-btm port 10 N4032-top port 2

I can see these group/ungroup messages whatever the ports are connected to N4032 active member. If I reload the active member then messages start appearing for other port. I can see the same effect on both force10 switches.

27 Posts

August 13th, 2014 11:00

N4032s are working on 6.1.0.6 version.

It seems like there are some packet drops on N4032 (other N4032 Transmitted packets dropped > MTU: ............8) but haven’t seen any discarded packets on S4810s.

As these switches are on the lab and nothing is connected to them, I can’t see any performance degradation etc…

N4032-btm#show interfaces counters port-channel 20

 Ch        InTotalPkts      InUcastPkts      InMcastPkts      InBcastPkts

--------- ---------------- ---------------- ---------------- ----------------

Po20                170278            61896           108373                9

 Ch        OutTotalPkts     OutUcastPkts     OutMcastPkts     OutBcastPkts

--------- ---------------- ---------------- ---------------- ----------------

Po20                158205            12669           132594            12942

Alignment Errors: ............................. 0

FCS Errors: ................................... 0

Single Collision Frames: ...................... 0

Multiple Collision Frames: .................... 0

Late Collisions: .............................. 0

Excessive Collisions: ......................... 0

Received packets dropped > MTU: ............... 0

Transmitted packets dropped > MTU: ............ 2556

Internal MAC Rx Errors: ....................... 0

Received Pause Frames: ........................ 0

Transmitted Pause Frames: ..................... 0

s4810-top#sh interfaces port-channel 11

Port-channel 11 is up, line protocol is up

Created by LACP protocol

Hardware address is 00:01:e8:8b:e9:80, Current address is 00:01:e8:8b:e9:80

Interface index is 1090977803

Minimum number of links to bring Port-channel up is 1

Internet address is not set

Mode of IPv4 Address Assignment : NONE

DHCP Client-ID :0001e88be980

MTU 1554 bytes, IP MTU 1500 bytes

LineSpeed 20000 Mbit

Members in this channel:  Te 0/10(U) Te 0/11(U)

ARP type: ARPA, ARP Timeout 04:00:00

Last clearing of "show interface" counters 6d11h49m

Queueing strategy: fifo

Input Statistics:

    229966 packets, 32927805 bytes

    440 64-byte pkts, 58266 over 64-byte pkts, 164844 over 127-byte pkts

    3690 over 255-byte pkts, 1 over 511-byte pkts, 2725 over 1023-byte pkts

    180366 Multicasts, 35960 Broadcasts

    0 runts, 0 giants, 0 throttles

    0 CRC, 0 overrun, 0 discarded

Output Statistics:

    222888 packets, 27793321 bytes, 0 underruns

    81225 64-byte pkts, 46654 over 64-byte pkts, 69174 over 127-byte pkts

    25782 over 255-byte pkts, 27 over 511-byte pkts, 26 over 1023-byte pkts

    158999 Multicasts, 19 Broadcasts, 63870 Unicasts

    0 throttles, 0 discarded, 0 collisions, 0 wreddrops

Rate info (interval 299 seconds):

    Input 00.00 Mbits/sec,          1 packets/sec, 0.00% of line-rate

    Output 00.00 Mbits/sec,          2 packets/sec, 0.00% of line-rate

Time since last interface status change: 21:23:56

27 Posts

August 13th, 2014 22:00

I ran below command but doesn't provide as much info as N4032. Hope it gives some idea...

s4810-top#sh lacp 11 counters

----------------------------------------------------------------------

             LACP PDU            Marker PDU         Unknown  Illegal

Port     Xmit       Recv       Xmit       Recv       Pkts Rx  Pkts Rx

----------------------------------------------------------------------

Te 0/0   45316      50095      0          0          0        0

Te 0/1   16500      476380     0          0          0        0

Te 0/10  4533       122569     0          0          0        0

Te 0/11  61067      77965      0          0          0        0

As soon as I've applied lacp long-timeout command then no new (group/ungroup) messages appears on the console. It is good to see that it is resolved however I have couple of questions regarding this issue,

1- I didn't come across "long-timeout" command on any vlt and mlag config docs. So what does it do? Do we need to run on all the vlt-mlag mesh configurations?

2-Also is there any trade off or pros/cons that we are going to get using long-timeout on lacp trunk?  

3- Afaik, following two commands "sh interfaces port-channel 11 brief , show vpc brief" are being used for mlag and vlt verification. So do you have any comments why we didn't see any lacp issue on these commands' output?

thanks,

No Events found!

Top