An 'alarm storm' - many alarms appearing across multiple cells at a single site within a short window - is a classic indicator that a shared resource serving all of those cells has failed. Multiple cells at a site commonly depend on the same System Module (baseband), the same transport connection back to the aggregation/core network, the same power supply, or the same synchronization source; a failure in any one of these shared elements can simultaneously impact every dependent cell, producing a burst of correlated alarms.
Recognizing this pattern allows the NOC engineer to search for a single common-cause fault - for example, a transport outage or System Module failure - rather than treating each cell's alarms as independent issues requiring separate investigation. It is extremely unlikely that multiple independent hardware faults occur simultaneously by coincidence, and such patterns should never be dismissed as a routine NetAct bug or simply ignored.

A threshold crossing alert (TCA) is generated when a monitored performance counter or KPI - such as PRB utilization, processor load, or a quality metric - crosses an operator-configured threshold value, providing visibility into a performance-related condition that may or may not correspond to an actual hardware fault.
For example, sustained high PRB utilization on a cell might trigger a TCA flagging a capacity concern, prompting capacity planning or optimization actions, even though no equipment has failed. This differs fundamentally from a standard equipment fault alarm, which is raised in direct response to a detected hardware or communication failure condition. TCAs are configurable by the operator, allowing thresholds to be tuned to the network's specific performance targets and capacity planning needs, and they are not limited to software license conditions, nor are they fixed, unconfigurable values.

Investigating a degradation in E-RAB or 5G QoS flow setup success rate for a specific gNB is most effectively approached by correlating multiple data sources covering the affected time period: checking active alarms on that gNB (hardware, transport, or signaling faults), reviewing the configuration management audit trail for any recent parameter change that coincided with the degradation, and examining relevant performance counters such as resource availability/congestion indicators, transport interface statistics, and signaling success/failure counts. This multi-source correlation identifies whether the degradation stems from a hardware fault, a configuration error, a capacity constraint, or a transport issue, enabling targeted remediation. Reviewing only high-level summary reports lacks the technical granularity required. Restarting NetAct is a drastic, unfocused action unrelated to a single gNB's issue and risks disrupting network-wide monitoring. Assuming the cause is always subscriber devices, without first checking network-side data, skips the most relevant and controllable areas of investigation.

When isolating a fault between baseband processing units and RF modules, checking the status and alarm details of the interconnecting interface - typically a CPRI or eCPRI link in AirScale architectures - between the System Module and the Radio unit is a logical first step. Alarms or abnormal status on this link can quickly reveal whether the problem lies in the connection itself (such as a fiber/cable fault or configuration mismatch) or within one of the connected components. This follows the principle of checking interfaces and boundaries before assuming a specific component has failed, avoiding unnecessary or premature hardware replacement. Replacing the antenna skips this diagnostic step entirely and risks an unwarranted field action.
Restarting NetAct is a network-wide action inappropriate for a single-site fault. Increasing transmit power to
'test' a suspected hardware issue is not a valid diagnostic method and could create interference.

Ping-pong handovers occur when a UE repeatedly hands over back and forth between two cells, often because the signal levels of both cells are very close to each other near the cell boundary, causing the handover trigger condition to be met in alternating directions in quick succession. Handover events (such as an A3 event, where a neighbor's signal becomes a defined offset better than the serving cell) include configurable hysteresis and time-to-trigger parameters. Increasing the hysteresis margin requires a more substantial and sustained signal difference before a handover is triggered, while increasing the time-to- trigger requires the condition to be met for a longer continuous duration before the measurement report is sent. Both adjustments reduce sensitivity to short-term signal fluctuations, directly addressing ping-ponging.
Disabling the neighbor relation would prevent legitimate handovers entirely, harming mobility performance; changing PCI values or bandwidth does not address the underlying signal-level trigger sensitivity that causes ping-pong behavior.