Skip to main content
The LARK is designed to run with minimal ongoing intervention. Once it is deployed and configured, most of what it does — capturing traffic, processing data, evaluating detection rules, running AI inference — happens continuously and automatically without any action on your part. That said, a small set of maintenance tasks benefit from regular attention to keep the appliance healthy and ensure that data is available when you need it. This page describes what those tasks are and how to think about them.

​
What Requires Regular Attention

Routine maintenance for the LARK consists of four areas. None of these require physical access to the appliance under normal operating conditions, and all are performed through the admin interface or the baseboard management controller (BMC).

Storage Utilization

Onboard storage is finite. Monitoring utilization regularly — and keeping it below recommended thresholds — prevents the rollover behavior that discards older data. If you have data offloading configured, verify that offload jobs are completing successfully and keeping pace with data accumulation.

System Updates

LARK software updates are applied through the admin interface. Staying current ensures you have the latest detection content, protocol decoders, and security fixes. Check for updates periodically and apply them during planned maintenance windows.

Planned Shutdowns

When you need to power down the appliance — for transport, scheduled maintenance, or site work — always use the orderly shutdown procedure through the admin interface or BMC. Hard power cycles risk leaving data writes in an inconsistent state and should be avoided except in emergencies.

Storage Recommendations

Beyond simply monitoring utilization, there are best practices around offload scheduling, understanding rollover behavior, and setting thresholds that help you get the most out of the available capacity. Following these recommendations reduces the risk of unexpected data loss from storage saturation.

​
What the System Handles Automatically

Many aspects of LARK operation are self-managing, and you do not need to intervene:
  • Data ingestion and processing run continuously at all times the appliance is powered on and monitoring traffic.
  • Detection rule evaluation happens inline, in real time, without a separate update or trigger step beyond loading new rule sets through the admin interface.
  • Storage rollover happens automatically when any storage path reaches capacity, preserving the most recent data without requiring manual cleanup.
  • Service health monitoring is built into the appliance — the admin interface surfaces service status, and the BMC can alert on hardware-level conditions independently of the OS.
The LARK has no rotating media or user-serviceable internal hardware components under normal operation. Physical maintenance — such as cleaning external connectors or reseating modules designed for transport — is limited to the components described in the hardware handling documentation.

​
Maintenance Task Reference

For step-by-step guidance on each maintenance area, refer to the dedicated pages in this section:
Establish a lightweight maintenance rhythm — check storage utilization and offload job status at a regular interval that matches your deployment’s traffic volume. In high-throughput environments, daily checks are reasonable; in low-traffic deployments, weekly may be sufficient.