LAN Switching Issues intermittently

[manniparashar]

Hello

I am facing some unusual issues since last week in my network.

I had upgraded to Cisco Catalyst 6807 XL as Core switch, Some of my servers misbehave intermittently and Users facing unusual drop from Servers and takes much time to access the Database and  tables from servers. 

When I changed the Network port from existing to another It works fine means Drops were not there as much. But later on another server misbehaves same as older one. After changing network port some times it works and sometimes not. 

Now, Presently We are facing intermittently drops in between from various servers but not able to resolve the issue.

Network Switch at various locations are stable and not having any issues even we are not getting drops from Network devices. 

Please help me out.

[Mark Malone]

hi

If the only thing that has changed since the upgrade is software I would roll back to stabilise the network to the software version that was working , at least then you can bring network back to normal and then put a test window in to see whats happening when you upgrade to the new version why some servers are becoming intermittent

are there any indications of issues in the logs and spanning tree alerts , cpu issues or mem , anything that could indicate a problem on the core switch

what about the interfaces are  they showing errors incrementing or are they clean , are the servers directly connected to the switch or through an access switch

Is it in vss mode or standalone core switch

[manniparashar]

Thanks Sir

We had a 4507 Switch as a Core gateway earlier and parallel we implemented 6807 switch. We are using the 6807 in parallel since last 8 Months. Now we only shifted the configuration part and dismantle the 4507 Switch from our network. 

Only 1 Vlan mismatch error is there in Logs of switch and no STP alerts, CPU and memory utilization are normal and not any other issue is indicated on switch.

Servers are connected via Server Farm switches i.e. 6800 IA and that too is connected with Core switch. 

Only 1 VLAN showing errors including giants (457 ), runts (45000)  and crc in Lacs.

This is standalone switch only.

Please help.

[dperezoquendo]

What are the connections between your server farm and core switch? Are they 10G fiber? Are there any errors on those interconnects?

Are you using QoS of any sort?

Also, what SFP modules are using? I once ran into an issue where the modules used on our core switch appeared to overheat. They were really hot to the touch and when left out to cool and then reinserted, it worked again.. for a few minutes.

As Mark mentioned, I would also check the software version you're on. You want to at least try to be on the Cisco suggested version, which appears to be 15.3.1-SY2(ED).

[manniparashar]

Thanks for reply Sir.

Server farm switches connected with Core switch via 10 G Fiber SFP Module.

No errors are there on any Ten GigaEthernet port which are connected to server farm switches. 

No QoS is configured on switch.

Temperature of core switch is stable and no heating is there on SFP Module.

Presently we are on version 15.2.1. 

Please suggest..

[Leo Laohoo]

Can you post the complete output to the following commands: 

sh interface Ten<BLAH>; and

sh counter interface Ten <BLAH>