This episode discusses centralized network management: the SNMP manager and agent concepts, SNMPv2c read-only community string configuration and the encryption advantages of SNMPv3, an introduction to AAA authentication authorization accounting with TACACS+ and RADIUS, and hardening Cisco devices by disabling unused services.

A network with dozens of routers and switches cannot possibly be monitored one by one through a console. Episode 19 discusses the final layer of operational maturity: how to monitor all devices from a single point, manage user access centrally, and close services that should not be open.
We discuss SNMP for monitoring, AAA with TACACS+ and RADIUS for centralized authentication, then close with hardening practices for Cisco devices. This material bridges your lab with professional network operations standards.
SNMP has two roles: an agent is the software inside a network device that provides data, and a manager is the server (NMS) that requests and displays that data. The manager sends queries, and the agent answers with metric values such as CPU, memory, and interface status.
Configure SNMPv2c on the router so it can be read by an NMS:
configure terminal
snmp-server community public RO
snmp-server host 192.168.10.100 version 2c public
exitsnmp-server community public RO defines a community string with
read-only access — the NMS can read but not change the configuration. The
second command directs traps and informs to the NMS address.
SNMPv2c sends community strings in plaintext, so anyone watching the packets could hijack access. SNMPv3 adds full encryption and authentication:
configure terminal
snmp-server group NMSGROUP v3 priv
snmp-server user monitor NMSGROUP v3 auth sha Pn3tLab2026 priv aes 256 Pn3tLab2026
exitsnmp-server user monitor NMSGROUP v3 auth sha ... priv aes 256 ...
creates an SNMPv3 user with SHA authentication and AES encryption. All SNMP
communication is now encrypted, eliminating the risk of reading community
strings off the wire.
AAA is a three-layer framework: Authentication verifies identity, Authorization determines access rights, and Accounting records activity. Two protocols carry it:
Basic AAA configuration with TACACS+:
configure terminal
aaa new-model
tacacs server TACACS-SRV
address ipv4 192.168.10.100
key Pn3tLab2026
exit
aaa authentication login default group tacacs+ local
aaa authorization exec default group tacacs+ localaaa new-model enables AAA, then login and exec authorization are directed to
the TACACS+ server with a fallback to the local database. If the server dies,
administrators can still log in via local.
A router's attack surface is as large as the number of open services. Turn off the services you do not use:
configure terminal
no ip http server
no ip http secure-server
no cdp run
exitno ip http server and no ip http secure-server turn off the web
management interface, while no cdp run disables CDP (Cisco Discovery
Protocol), which sends device information to every neighbor.
The information collected by CDP is a goldmine for attackers on a local
network.
Besides HTTP and CDP, also audit no ip bootp server, no service tcp-small- servers, and disable Telnet with transport input ssh (episode 2).
Get into the habit of a hardening checklist: every open service must have a
clear reason, and the rest are closed.
Build one router and one Docker Linux server acting as both the NMS and the
TACACS+ server. Configure SNMP on the router, run snmpwalk from the NMS to
read OID values, then test AAA by logging in via TACACS+. This exercise
connects all of episode 19's concepts into a single complete operational flow.
Key takeaways:
RO community can only read; SNMPv3 adds encryption.no ip http server and no cdp run close unused services.In the next episode, episode 20, we summarize the entire series in one big project: complex enterprise lab & comprehensive troubleshooting — designing a complete multi-site topology with VLAN trunking PortFast, core switches with SVI EtherChannel and HSRP, a WAN edge with multi-area OSPF and eBGP to an ISP, a branch GRE over IPsec VPN, and a layer-by-layer troubleshooting methodology with the CLI verification commands you must master.