Template:FAQ:Applications 05

There are a number of possible causes of this.

The most likely are the agent access control settings (who is allowed access by the agent itself), or firewall/packet filtering settings (who is allowed access by the underlying operating system).

A fuller list of possible causes (with indications of how to check for each) is as follows:

• is the machine you are querying up and running?
  (Does it respond to 'ping' or similar requests?)
• is there an SNMP agent running on it?
  (Run 'ps -ef | grep snmp' or 'netstat -an | grep 161')
• are the requests arriving, or being blocked (e.g. by a firewall)?
  (Restart the agent using 'snmpd -f -Le -d' and see if it shows the incoming packet dumps)
• is the agent simply taking a long time to respond?
  (The 'snmpd -f -Le -d' command should show a series of incoming PDUs, followed eventually by the outgoing PDUs.
  Try the request again with a long timeout value, e.g. 'snmpcmd -t 120 ....')
• do the agent's control settings allow this request?
  (The 'snmpd -f -Le -d' command will show a series of incoming PDUs with no corresponding outgoing PDUs)

If the agent is not configured to allow access for a particular community, then no error response will be returned. The Net-SNMP tools will retry the request a number of times, before reporting a timeout error.

If the agent is configured to allow partial access for a given community, then requests that fall outside this authorised access will result in an error response. (SNMP agents can be very fussy over who they talk to!)

See the entries on access control in the AGENT section for how to configure the Net-SNMP agent to allow suitable access. For other vendors' agents, you will need to consult the relevant documentation.