input errors on port channel cisco switch to blade center

[fh497a]

hello,

connecting from a cisco 4506 to a cisco switch on a bladecenter hs21. The delay is set to 10usec on all ports and port channels.

here is the config on the 4506

interface Port-channel2
description BLADECENTER SW2
switchport
switchport trunk encapsulation dot1q
switchport trunk native vlan 3
switchport trunk allowed vlan 1-3,7,14,20
switchport mode trunk

interface GigabitEthernet5/22
description BLADECENTER-sw2-p17 Truck Port
switchport trunk encapsulation dot1q
switchport trunk native vlan 3
switchport trunk allowed vlan 1-3,7,14,20
switchport mode trunk
channel-group 1 mode on
end


bladecenter switch

interface Port-channel2
switchport trunk native vlan 3
switchport trunk allowed vlan 1-3,7,14,20
switchport mode trunk
delay 1
!

interface GigabitEthernet0/17
description extern1
switchport trunk native vlan 3
switchport trunk allowed vlan 1-3,7,14,20
switchport mode trunk
macro description cisco-switch
channel-group 2 mode on

this is a show int po1 on the 4506

Port-channel2 is up, line protocol is up (connected)
Hardware is EtherChannel, address is 000e.8329.1334 (bia 000e.8329.1334)
Description: BLADECENTER SW2
MTU 1500 bytes, BW 2000000 Kbit, DLY 10 usec,
reliability 255/255, txload 1/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 1000Mb/s, media type is N/A
input flow-control is off, output flow-control is unsupported
Members in this channel: Gi5/21 Gi5/23
ARP type: ARPA, ARP Timeout 04:00:00
Last input 10:02:03, output never, output hang never
Last clearing of "show interface" counters never
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 2112000 bits/sec, 461 packets/sec
5 minute output rate 1644000 bits/sec, 523 packets/sec
23480763 packets input, 13239031193 bytes, 0 no buffer
Received 191947 broadcasts (177535 multicast)
0 runts, 0 giants, 0 throttles
37 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored

 

[fh497a]

going through log shows im seeing the below. The bladecenter vtp status is set to transparent. Could this be causing the issues?

2. If I do change the bladecenter vtp client to server will this cause any loops?

*Jan 31 05:06:38 PST: %DTP-5-DOMAINMISMATCH: Unable to perform trunk negotiation
on port Gi5/24 because of VTP domain mismatch.

 

[fh497a]

i did remove the error by assiging the same VTP domain .. but stil getting input errors.

 

[leedsit]

Hi, If you only have 37 input errors, I would`nt worry about it. They may have occured when the Port channel was being established. I use roughtly the same setup and get a few errors on establishment. Rather than force the port channel on, if you use a negotiated setting you probably would`nt see any errors.

Lee.


LEEroy
CCNP,CCIP

 

[fh497a]

the errors continue to grow daily .. I had 37 errors after tranferring a 600mb file to the server.

which command do you suggest?

active Enable LACP unconditionally
auto Enable PAgP only if a PAgP device is detected
desirable Enable PAgP unconditionally
on Enable Etherchannel only
passive Enable LACP

 

[ADB100]

Are you seeing errors on the physical ports? The delay parameter is not introducing a delay, it is used by routing protcols in calculating the routing table.

http://www.cisco.com/en/US/tech/tk365/technologies_white_paper09186a0080094cb7.shtml

My recommended configuration would be to force the channel on, disable DTP, use a unique unused Native VLAN and also tune the load-balancing algorithm so it utilises each port the best.

port-channel load-balance src-dst-ip (or whatever suits your traffic best)
!
interface GigabitEthernet5/22
 switchport trunk encapsulation dot1q
 switchport trunk native vlan 4000
 switchport trunk allowed vlan 1-3,7,14,20
 switchport mode trunk
 switchport nonegotiate
 channel-group 1 mode on

Also VLAN 1 has special significance on Cisco trunks so you shouldn't really use it for user data, however I understand that it is probably a legacy thing.

Andy

 

[baddos]

I would replace your patch cables first. If your are using some cross connects too, verify that cabling is good as well.

 

[fh497a]

yes im showing input errors on the physical ports. Will need to review the load balancing algorithms.

But we are using this for (if you can recommend an algorithm)

esx and physical servers that host sql, sharepoint, fileserver, webservers ....

Will the below have any effect on anything else. I didn't do the intial bladecenter switch config. Do I have to actually setup vlan 4000 to be routable? or not? I wouldn't think so since we are just using it to pass traffic.

switchport trunk native vlan 4000

 

[ADB100]

The Native VLAN must exist and it should be not used anywhere else and should just be on the P2P link between these two switches. It definitely should NOT have a Layer-3 SVI.
The load-balancing algorithm needs to be tuned to reflect the traffic - i.e. if all your Blade Centre Servers are on different VLANs to clients then dst-mac on the Blade-Centre switch doesn't give you much load-balancing since the destination MAC will always be the same (the SVI on the Cat 4500). I usually go for src-dst-ip as these will usually vary?

I'd definitely swap out the physical cabling, maybe GBICs as well if these are used? Other than that I'd make sure you are running the latest IOS versions for the blade switch and the 4500.

Andy

 

[fh497a]

Bladecenter servers are distributed between two vlans.
I've already changed the physical cabling and ports.
Will try upgrading the IOS on both sides