Monitoring a switch with SNMPv3

On the switch

In the switch’s management settings, turn on SNMP v3 for the TarleyNode server’s address, with a user that has both authentication and privacy (authPriv). Note the user name, both passwords and the two protocols you chose, because TarleyNode must match them exactly.

Save the login in TarleyNode

1. Open Discovery › Credential profiles and choose Add.
2. Give it a Profile name, such as Core network · Read only, and choose the Credential type SNMP v3 · authPriv (recommended).
3. Enter the Username, the Authentication password and the Privacy password. Each password needs at least 8 characters.
4. Under Authentication, pick the same protocol as the switch: SHA-1, SHA-224, SHA-256 (the default), SHA-384, SHA-512 or MD5. Under Privacy, pick AES-128 (the default), AES-192, AES-256 or DES.
5. Save. The login is encrypted at rest on your server.
6. Administrators can also create a login from a device’s page: under Logins, choose Assign (or Change), choose Create a new login if saved logins are listed, fill in the form and choose Create and assign.

Add the switch

1. Make sure the switch’s address is inside a site’s ranges under Discovery › Sites & network scopes.
2. Either choose Scan now and Add the switch from the results, or open Devices › Add device and enter its name, address and site. A ping check is added straight away.
3. On the switch’s page, under Logins, assign the SNMP login. To give every switch the site finds the same login, set it as the site’s SNMP login instead.

What TarleyNode then monitors

• Every active port, added automatically: traffic, errors and status.
• Hardware sensors, found and added automatically: temperature, fans, power supplies and PoE. Remove any you do not want.
• CPU, memory and uptime, with restart detection from the device’s uptime, even between checks.
• The maker’s template, where TarleyNode recognises the maker, shown on the Sensors card.
• The port map, which shows which device is plugged into which port, and the network map drawn from LLDP and CDP neighbours.

Traps and syslog

The same SNMP v3 user is accepted for SNMPv3 traps from the devices it monitors. Send the switch’s traps to the TarleyNode server: syslog on UDP 514 and SNMP traps on UDP 162 arrive under Events, where event rules can turn messages into incidents. Events shows the last 3 months, and Include older entries searches everything that is kept.

Tune it

1. To check a core switch more often, open Edit device and set Check every, from 15 s to 1 h. Every check on the device moves to that interval, including checks added later.
2. Under Rules, add thresholds on any value, on top of the default rule that down means an incident.
3. If the switch can export flows, open its page and choose Send flows to TarleyNode for the commands to paste into it.

If it does not work

• Check the switch is inside the site’s ranges. TarleyNode only ever contacts addresses inside them.
• If the check reports The SNMP request failed., the user, passwords or protocols do not match the switch. The console’s own example: a FortiGate uses SHA-1 or SHA-256 with AES.
• If it reports The device cannot be reached from where it is monitored., check the route and any firewall between the switch and the TarleyNode server or connected site.

Switches at other locations

For a switch this server cannot reach, connect that location under Connected sites › Connect a site, then choose the connected site as where the switch is monitored from when you add it. The connected site uses the same SNMP login and keeps readings on disk if its link drops.

Related

• SNMP monitoring
• Network map from LLDP and CDP
• Moving from PRTG to TarleyNode
• All features

TarleyNode is offered by invitation while it is new. Request a demo to try it on your own network.