Unsolved

1 Rookie

 • 

10 Posts

1133

February 24th, 2020 11:00

Oversize Packets Between New and Old Dell Switches Causing Packet Drops

Hello everyone,

I am having a bit of trouble in my network and I was not able to find the root cause. I've been investigating for several weeks, and the only problem I fond are some oversized packets on some LAGs and I am not sure if that is the actual problem.

We recently replaced some old switches (Dell PowerConnect 2848) with new ones (Dell X1052P). Due to inability of configuring the management VLAN on other than VLAN 1 on old switches, we have configured the new ones with the management in the same VLAN 1. We also had to create General ports in between switches to be able to transport the other VLANS as tagged while the management VLAN untagged. 

After the replacement has been done, some strange things started to happen in the network, and I was not able to address or troubleshoot them yet. Other equipments in the management network started to drop packets unexpectedly without changing any configuration on them. I've started to monitor the switches and I saw some oversized packets on some switches (on LAGs connecting new switches with old ones). 

What did I do to troubleshoot so far:


- I've been investigating for layer 2 loops and didn't found any.

- I've checked physical cables or deteriorated switch ports and there aren't any.

- I've checked the configurations multiple times and didn't find anything suspicious or out of common.

 -- Checked the VLANs configurations

 -- Checked the LAGs and ports configurations

- I've checked the CPU on the switches and it runs at about 20-30% load.

- I've checked the logs on the new switches and they don't complain about anything

 

I am not sure if the layer 2 or the switches themselves are responsible for the network behaviour and poor performance, but it all started when we replaced some of the old ones. Beside the oversized packets there are a lot of drops at layer 3. Pinging devices in the same VLAN has a success rate of about 80%, while it use to be 100%.

I've attached a picture of the LAG monitoring of an old switch and a new one. The oversized packets are seen only at the old one end.


X1052PX1052PPowerConnect 2848PowerConnect 2848

 

This is how the interfaces and port-channels are configured:


interface tengigabitethernet1/0/1
channel-group 3 mode on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface tengigabitethernet1/0/2
channel-group 3 mode on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface tengigabitethernet1/0/3
channel-group 1 mode on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface tengigabitethernet1/0/4
channel-group 1 mode on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface Port-channel1
flowcontrol on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface Port-channel2
flowcontrol on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface Port-channel3
flowcontrol on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!
interface Port-channel4
flowcontrol on
switchport mode general
switchport general allowed vlan add 2-5,7 tagged
switchport general allowed vlan add 1 untagged
!














































 

 

 

The firmware is up to date. I am not sure if there is a mismatch or a compatibility problem between the new switches or the old ones.

I am not sure if the problem was well explained, but hopefully someone can tell me if there is anything wrong?

Thank you!



Moderator

 • 

9.4K Posts

February 25th, 2020 09:00

Hi,

The number of oversized packets isn’t that high, so that shouldn’t be the problem. Is it just specific devices that are getting dropped packets or is it across all of them? You may want to try using wireshark to see if it provides more information. You have done a lot of troubleshooting already.

1 Rookie

 • 

10 Posts

February 26th, 2020 08:00

I have also configured a packet capture solution in the network to check what is going on, but there is nothing more than usual. I was looking specifically at layer two, but there are just usual ARPs, DHCP and a few CDPs frames.

 

Moderator

 • 

9.4K Posts

February 26th, 2020 09:00

In that case I am not sure what is causing the drops. 

1 Rookie

 • 

10 Posts

February 27th, 2020 00:00

Do you think there could be an incompatibility between the old switches and the new ones? 

That would be the most difficult troubleshoot step to do in order to narrow the problem down, and I don't even know where to start.

Also, I know the protocols should work, because that's why we have standards for them, but is there any chance that the new switches uses more sophisticated signalling methods that the old ones don't understand?

Thanks

Moderator

 • 

9.4K Posts

February 27th, 2020 06:00

I think the best option is for you to call phone support and see if they have any additional ideas.

No Events found!

Top