Ethernet Switches & Routers

Reply
Occasional Contributor
Posts: 11
Registered: ‎09-02-2013

ICX6450 Stack Problem with HP NCU Teaming

[ Edited ]

Hi,

 

We are facing weired problem in the stack of 7 swicthes. We have tried all the possible options to isolate the issue permanently and we got failed.

 

We have 7 units of ICX6450-48 which we have connected as stack in Ring Toplogy. The Stack connections are connected each other with 10G optics. We have the LACP dynamic configuration with our HP storage ( we have bundled 1Gbps x 8 ethernet cables with HP NCU utility).  Everything seems working fine from past two years and suddenly users are expierenced the slowness storage access problem in the network.

 

We have contacted HP and they couldn't able to find any missmatch or hardware fault in the storage and we have forced them to change the NIC relacement. Even after the problem remains as it is. When the problem starts we see the "frag aged error " in the switch and sometimes the stack ports are widely incrementing the "input errors". There is no pattern of the problem and this will be intermittent issue. When we copy 1GB file from storage, the users are getting 345 to 500KBps speed over 1Gbps ethernet card at user end and somtimes they are getting intermittent RTO.

 

Best, Surendra

 

 

 

We have configured below on switch level for LACP.

 

!

interface ethernet 1/1/1
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/2
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/3
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/4
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/5
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/6
 no spanning-tree
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/7
 link-aggregate configure key 10000
 link-aggregate active
!
interface ethernet 1/1/8
 link-aggregate configure key 10000
 link-aggregate active

 

 

Tagged on VLAN 10 to pass the traffic

 

vlan 10 by port

tagged ethe 1/1/1 to 1/1/8

 

The below error we are getting from switch when we face slowness  file transfer problem

 

telnet@BGLRECNSXTX001#show mem   stat stack-ports

Port            In Packets          Out Packets       In Errors      Out Errors
1/2/1            621575963           1317314499               0               0
1/2/3            752681167           1571764475               0               0
2/2/1            423250680            199482875               1               0
2/2/3             17829332             26312261               0               0
3/2/1            917617467            441619789               0               0
3/2/3            199481873            423249537               0               0
4/2/1            446820347            886220701               0               0
4/2/3           1316891445            621387184              20               0
5/2/1           1571335093            752601398              69               0
5/2/3            441616744            917633252               0               0
6/2/1             26273373             17808059               0               0
6/2/3            327125762            164770229           22957               0
7/2/1            164000681            327758561               0               0
7/2/3            880796708            441437262               9               0

TOTAL           8107296635           8109360082           23056               0

 

 

telnet@BGLRECNSXTX001#show stat stack-ports             ck ipc
V16, G3, Recv: SkP0: 235646, P1: 223496, MAIL: 14448554, sum: 14907696, t=30232.5
Message types have callbacks:
    1 : Reliable IPC message 2 : Reliable IPC atomic  4 : fragmentation, jumbo
    5 : probe by mailbox     6 : rel-mailbox          7 : test ipc             
    8 : disable keep-alive   9 : register cache       10: ipc dnld stk         
    11: chassis operation    12: ipc stk boot         13: Rconsole IPC message
    14: auth msg             15: ipc erase flash      16: unconfigure          
    17: ipc stk boot         18: ss set               19: sFlow IPC message    
    21: SYNC download reques 23: SYNC download 1 spec 28: SYNC client hello    
    30: SYNC dy chg error    32: active-uprintf       33: test auth msg        
    34: probe KA             39: unrel-mailbox        40: trunk-probe          
    41: mail-nbr             
Send message types:
    [1]=188616,       [4]=2710,         [5]=357599,       [9]=1813893,      
    [11]=2996,        [20]=17344,       [22]=38775,       [25]=101,         
    [26]=45996,       [29]=21996,       [34]=117508,      
Recv message types:
    [1]=161094,              [4]=17522,               [5]=176994, 164739,      
    [9]=14191568,            [11]=48983,              [21]=38775,              
    [30]=2468,               [34]=58652, 58757,       

Statistics:
    send pkt num          :   2490016,    recv pkt num          :  14907696,    
    send msg num          :   2490026,    recv msg num          :  14910796,    
    send frag pkt num     :      1355,    recv frag pkt num     :      8756,    
    pkt buf alloc         :   2490016,    

    Reliable-mail            send  success recveive   duplic
    target ID                   0        0        0        0
    target MAC                  0        0        0        0
    unrel target ID             0                 0         
There is 1 current jumbo IPC session

Possible errors:
    *** frag aged         :         5,

 

Occasional Contributor
Posts: 11
Registered: ‎09-02-2013

Re: ICX6450 Stack Problem with HP NCU Teaming

Hi All,

 

The problem has been resolved by changing the one of switch in stack.

 

Best, Surendra

Join the Community

Get quick and easy access to valuable resource designed to help you manage your Brocade Network.

vADC is now Pulse Secure
Download FREE NVMe eBook