snmp error but snmpwalk is working fine

[mitro]

Hello,

I have read lots of documentation and how to but I don't figure how to solve my problem:
I need to poll informations (Number of packets accepted) from Checkpoint software:

From cacti host, I do the following (of course replacing $communityString and $ipAddress with real ones):

-bash-3.00$ snmpget -O vt -c $communityString -v 1 -t 500 $ipAddress:260 .1.3.6.1.4.1.2620.1.1.4.0
INTEGER: 153251022

-bash-3.00$ snmpwalk -c $communityString -v 1 -t 500 $ipAddress:260 .1.3.6.1.4.1.2620.1.1.4.0
SNMPv2-SMI::enterprises.2620.1.1.4.0 = INTEGER: 153357718

Everything is working fine, since I get the expected results.
Note that the snmp port is not standard (260 instead of 161, the snmp daemon is the Checkpoint one so that is ok).

When I add a host / device in Cacti and set up the right settings (ip address, snmp version, community), Cacti displays the well known "snmp error" message.
I am totally confused since *everything* is working fine in command line / shell.

Please could you help me to solve this issue, or point to an existing documentation which could help me to solve it?

Thanks a lot.

[BSOD2600]

Does a snmpwalk of that hosts uptime work?

[mitro]

Does a snmpwalk of that hosts uptime work?

Excuse me but could you be more precise? I don't understand what do you mean and what do you want me to do.
Snmpwalk is working fine from cacti host to the destination host.

Again, thanks for your help.

[BSOD2600]

Do something like:

snmpwalk -c $communityString -v 1 $ipAddress:260 system

Part of cacti's test to determin if a device is up or down is if it can get the uptime/sytem info. If it cannot, then it'll consider the device down.

What uptime detection method do you have set in cacti's settings?

[mitro]

Code: Select all

-bash-3.00$ snmpwalk -c $communityString -v 1 $ipAddress:260 system
-bash-3.00$

And the uptime detection method is set to snmp.
Then this is the explanation... The system parameter does not send back any value so Cacti assumes the host is down.

Do you know if there is a way to fix this?
I could change the Downed Host Detection from snmp to Ping for example?
(Cacti monitors 200 hosts here, and I have to be careful with modifications).

[BSOD2600]

Then this is the explanation... The system parameter does not send back any value so Cacti assumes the host is down.

That'd be my assumption too. Really should contact your manufacturer and get them to properly implement that, since it's fairly common practice for SNMP enabled devices.

Do you know if there is a way to fix this?
I could change the Downed Host Detection from snmp to Ping for example?
(Cacti monitors 200 hosts here, and I have to be careful with modifications).

Changing the downed host detection as you've suggested is probably the best option for this SNMP enabled device (otherwise, I'd tell you to remove the snmp community info and cacti would always poll it).

[gandalf]

Changing the downed host detection as you've suggested is probably the best option for this SNMP enabled device (otherwise, I'd tell you to remove the snmp community info and cacti would always poll it).

Nope, AFAIK. Cacti will detect it as up but will lack the community string needed to access data via SNMP. This solution buys you another problem, sorry
Reinhard

[BSOD2600]

Nope, AFAIK. Cacti will detect it as up but will lack the community string needed to access data via SNMP. This solution buys you another problem

Err, right. Thats what I was after