Figure 1 illustrates the flow of processing an SNMP request when the request is for data supported by the TCP/IP subagent.
Figure 1. Overview of SNMP support
The following list describes the sequence of events that occur from the time you issue an SNMP command until you receive the response:
- The user issues a NetView® SNMP command or a z/OS® UNIX snmp command.
- The command processor validates and encodes the request in a Protocol Data Unit (PDU), and sends it in a UDP packet to the SNMP agent. The SNMP agent can be on the same MVS™ image or on a different MVS image.
- The SNMP agent validates the request and, if necessary, sends it to an SNMP subagent. Requests for agent-oriented objects are handled by the agent and all others are handled by a subagent. To determine which objects are handled by the agent and which objects are handled by a subagent, see the Management Information Base (MIB) objects appendix in z/OS Communications Server: IP System Administrator's Commands.
- The agent sends the response to the originator of the request. The command processor displays the response.
Tip: Although not shown in Figure 1, other subagents that
are shipped as part of z/OS Communications
Server, such as the OMPROUTE subagent, the Network SLAPM2 subagent,
and the TN3270E Telnet subagent, also communicate with the SNMP agent
using AF_UNIX socket calls from their own address spaces.
The SNMP agent and the SNMP subagents record trace information through the z/OS UNIX syslog daemon (syslogd) using the daemon facility. For more information about syslogd and specifying the daemon facility in the /etc/syslog.conf configuration file, see Logging of system messages.