cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
SKH
New Contributor

NetVanta 1560-24-740W CPU 100% All The Time

The switch seems to work but with the CPU stuck at 100% almost nothing gets through, DHCP appears to work intermittently, and so on. We're assuming we have misconfigured something but we haven't been able to figure out what. The CPU hits 100% and never calms down. We have it with as close to nothing connected as possible and CPU remains at 100% overnight. Two identical switches, connected together with 10Gbps SFP module or using a 1Gbps port, a single connection to the office network or connected directly to a laptop. Its been tested with and without devices connected, with and without SFP modules, connected to our SD-WAN sockets or connected to another non-ASE switch. No change.

VAN-SW-2362073# show system cpu status
Average load in 100 ms : 100%
Average load in 1 sec : 100%
Average load in 10 sec : 100%
 
Attached is the running config from the switch (hashed passphrases removed). Note the VLAN configs are necessary (1,2,3 & 100) but we're new to ASE and suspect we've made a mistake in there somewhere.
Tags (1)
0 Kudos
2 Replies

Re: NetVanta 1560-24-740W CPU 100% All The Time

SKH, thank you for providing the config. A couple of questions first off.

What version of firmware?
What sort of devices are typically going to be plugged in to the ports, for example Gig 1/1-22 ?
What is your goal with the QoS settings on each port?
SNMP polling can tax the CPU pretty heavy, have you tried turning of SNMP on the switch and/or limit what you are polling from your SNMP server to see if the CPU calms down?


If you have a ProCare plan, you can open a ticket and tech support can dig deeper into the switch to see what processes are consuming the most CPU cycles.
BMullar
New Contributor III

Re: NetVanta 1560-24-740W CPU 100% All The Time

SKH,

I have not used the NetVanta 1560 devices yet.  Yet I have the following questions:

Q1:  Are the two 1560 switches connected with two Trunk ports?  If yes, I do not see "aggregation group" command on the 10Gb interfaces, which are in Trunk mode.  So no LAG managing traffic over those two links for a potential Ethernet loop.

Q2:  Spanning Tree is off by default but you have "spanning-tree mst" and "spanning-tree aggregation" commands.  Is STP provisioned and running?  If yes, then it should block traffic on one of the 10Gb interfaces to prevent a loop.

Q3:  If the rest of the LAN is disconnected from these 1560s and the CPU remains at 100%, that is why I am thinking about an Ethernet loop.  Does the CPU stay at 100% if the rest of the LAN is disconnected?  If yes, again an Ethernet loop.  If no, then a device on the LAN is generating the traffic.

Q4:  Every port is provisioned at "switchport mode hybrid", but the default is "switchport mode access".  Why did you set them for hybrid which is "access" or "trunk".  Change back to "access" as a test.

 

Does the following display the same CPU utilization results?  I am wondering what is the trigger for the CPU to hit 100%

1560 $1...nothing connected.

WAN > Router > 1-Interface 1560 #1

WAN > Router > 1-Interface 1560 #1 > Laptop

WAN > Router > 1-Interface 1560 #1 > 1-Interface 1560 #2

WAN > Router > 1-Interface 1560 #1 > 1-Interface 1560 #2 > Laptop

WAN > Router > 1-Interface 1560 #1 > 1-Interface 1560 #2 > LAN

 

Hope this helps.

 

Regards,

BMULLAR