ProCurve 2810's Flapping and dropping packets to management ip address

[DanKingdon]

Hi All,

Looking for some help with an intermitent problem we are having. 

We have a site of approx 30 procurves 2810's 24 and 48 port with approx 800 network devices.  At the core are two "Distribution" sitches and a core switch.  The core switch has all the servers connected and the dist switches connect via fibre to the access switches. 

The core connects to both dist switches via a 4 port trunk and each dist switch connects to each other with a single dual port trunk.  Spanning has disabled this as expected. We have 1 default vlan that most clients connect to and two other vlans for wifi and other traffic.

The core switch (and the dist switches to a lesser extent) are dropping packets to the management ip address.  Every couple of mins or more the pings go through the roof and eventuall drop a couple of packets before recovering.  The switches are forwarding traffic fine during this time but we are starting to see some STP recalculations happeneing if either of the dist switches lose connection to the the core switch for long eneough.

The cpu usage is around 20-27% on the core switch, nothing is reported in the event log of any of the switches.  I have debugged STP and all seems fine. 

1.  Is this just the result of having such a large network all on the same VLAN?  ie overloading the switch management with to much traffic?

2.  How else can I troubleshoot what is causing the switches to drop packets?  Is my next step to run a packet capture with wireshark to inspect the traffic between the switches?

3. Are these switches just not cut out for this much use?  (We are currently looking to fit an E5406vl soon to replace all three of these switches)

Cheers