JNCIP-DC: Data Center Monitoring, Troubleshooting, and Operations

This domain tests whether you can localize a fault across underlay, overlay control plane, and data plane, plus day-to-day operations such as telemetry, logging, and rollback.

Layered troubleshooting order

1. Underlay: show bgp summary, show route, loopback-to-loopback pings, BFD state, interface error counters.
2. Overlay control plane: show route table bgp.evpn.0, show evpn database, show evpn ethernet-segment.
3. Data plane: show ethernet-switching table, show arp no-resolve, monitor traffic interface xe-0/0/1 no-resolve.
4. Platform: show chassis alarms, show log messages, interface statistics.

Operations

• Streaming telemetry through JTI/gNMI gives trend data; keep SNMP only for legacy collectors.
• Use commit history and rollback for change recovery, and request support info when opening a case.
• Apstra probes and anomalies provide fabric-wide health; centralize syslog for correlation.

Worked example

Host A on leaf1 cannot reach host B on leaf2 in the same VLAN. Underlay BGP is up and loopbacks ping, so the fault is above the underlay. On leaf2, show evpn database lists B's MAC. On leaf1 it does not. show route table bgp.evpn.0 on leaf1 shows the Type-2 route is present but not installed, which points to an import problem: the VNI route target on leaf1 does not match leaf2's export target. Fix protocols evpn vni-options vni ... vrf-target, then confirm the MAC appears in show evpn database and traffic flows. If the Type-2 route were absent entirely, check VTEP source reachability and VNI configuration instead.

Traps

• Debugging the data plane before confirming loopback reachability.
• Treating show evpn database as forwarding proof; it reflects control-plane state only.
• Missing MTU or fragmentation issues where large pings fail but small ones pass.
• Clearing or restarting protocols before capturing show output and logs for the case.