Network Management or Reliability
This domain tests design for manageability and resilience in Junos networks: OOB access, telemetry, configuration safety, and high-availability mechanisms.
Core rules
- Build out-of-band management with a dedicated VRF, console servers, and ACLs.
- Use SNMPv3 authPriv, syslog to redundant collectors, and authenticated NTP.
- Prefer NETCONF/YANG and Junos Telemetry Interface over polling for scale.
- Use commit confirmed, rollback, and configuration groups to reduce change risk.
- Design redundancy: GRES, NSR, BFD, LAG/MC-LAG, and dual Routing Engines where supported.
Worked example
A 200-device WAN needs 99.99% management availability and rapid fault isolation. Design: OOB VRF on all SRX and EX devices, dual syslog collectors, SNMPv3, NTP from internal stratum-2, telemetry to a collector, and BFD on WAN links. Reason: in-band management fails when the WAN fails; OOB preserves recovery. SNMPv3 avoids clear-text community strings. BFD detects sub-second failures so routing converges before users notice.
Traps
- In-band-only management and no console access.
- SNMPv2c and unauthenticated NTP.
- Relying on polling at thousands of interfaces.
- Enabling GRES or NSR without checking platform support.