Graph error because of SNMPwalk problem

[yoas]

Hi all,
i am newbie about cacti. I have a cacti that monitor my servers (solaris).
I have one problem because cacti can not show the graph well for one server. I have tried to ping it and running well. Then i tried to snmpwalk:

--------(at the last response)

Code: Select all

HOST-RESOURCES-MIB::hrDeviceIndex.1560 = INTEGER: 1560
HOST-RESOURCES-MIB::hrDeviceIndex.1561 = INTEGER: 1561
HOST-RESOURCES-MIB::hrDeviceIndex.1562 = INTEGER: 1562
Timeout: No Response from 10.22.126.31:33161

When i used other PRTG (such as infovista) it can running well even part of snmpwalk failed.
Have you a suggest what i should do for this case? How to reach graph even snmpwalk does not running well?

Thank you

[fmangeant]

Hi

what kind of device are you trying to graph ?

[yoas]

my device is unix solaris server. I have more than 5 servers like that, and run well. Just this one have a problem. Can you help me?
Thank you

[gandalf]

Which version of cacti? Which settings for Maximum OID's Per Get Request?
Reinhard

[yoas]

On my setting, The Maximum SNMP OID's Per SNMP Get Request = 60 (max.) and i use cacti version 0.8.6i. Can u help me... please...
thank you

[fmangeant]

Can you decrease this setting to 10 ?

[yoas]

can you explain to me waht the effect when i change this setting?
i must aware because beside solaris servers, i also have hundreds routers that monitored by cacti.
thank you

[gandalf]

This setting is only used with SNMP V2c and V3. It's a performance option. When using it, cacti will tray to instruct SNMP to sent that many OID requests in one single SNMP packet. Sounds good, eh? But unfortunately, some target devices don't like that.
That's why we decided to make a host-specific option out of it.
It may require some try and error to find the settings for a device, that does not like huge settings
Reinhard