Simple Network Management Protocol (SNMP)

SNMP works by sending messages, called Protocol Data Units (PDUs), to different SNMP compliant devices, called agents, store data about themselves in Management Information Bases (MIBs) and return this data to the SNMP requester's. SNMP originally designed to monitor the network devices like routers and switches. But now they are using to retrieve the information about the entire network infrastructure. System up time, CPU, disk and memory usage are few of them.

SNMP Architecture

The SNMP architecture consists of three elements:

1. Manager: Software runs on the system, designed to communicate and manage network devices which has agents.

2. Agent: Software stored on network devices like routers, switches etc. and provides information to the managers.

3. MIB: Just a text file, and MIBs data object values are used for the communication between managers and agents.

MIBs and OIDs Overview

SNMP uses an extensible design of the available information defined by MIBs. These MIBs can be described as a tree like structure of the management data of a device, which use a hierarchical name space known a s object identifiers (OID). Each OID defines a unique variable that can be read or set via SNMP.

SNMPTrap: Is a SNMP application that sends an SNMP notification (trap) to a network administrator.

Applications

SNMP is widely used by network administrators to monitor and manage network performance, detect and resolve network issues. Even you can collect the long term data.

Note:
The first two versions of the SNMP protocol (SNMPv1 and SNMPv2c) contained clear text passwords called community strings. SNMPv3/USM was later developed which include a User-Based Security Model, that allows network administrators to secure the protocol by defining the users with various security credentials.