Skip to main content
Hardware

TERRA Server (Wortmann)

Monitor the hardware of your TERRA and Wortmann servers via Redfish through the management card, from temperatures and fans to drives and hardware events.

The TERRA Server (Wortmann) check monitors hardware through the server's management card (BMC). It uses HTTPS and Redfish to read temperatures, fans, power supplies, voltages, memory and drive health, and hardware events, every 15 minutes by default. Available data depends on the management card and connected hardware.

Prerequisites

  • You need Monitoring Check Management and access to the device on which you configure the check.
  • An online octoja agent must be able to reach the management card over HTTPS. The BMC address is not the server operating system's address.
  • The BMC must support Redfish. Older TERRA servers with an Intel RMM4 module offering only IPMI cannot use this check.
  • Create a dedicated BMC account with the Operator or ReadOnly role and have its credentials ready.

Steps

  1. Under Devices, open the device with the agent that can reach the BMC.
  2. Choose Checks → Add Check → TERRA Server (Wortmann).
  3. Enter the management card's hostname or IP address under Host, without a URL path. Fill in BMC user, BMC password and, if necessary, a different HTTPS port.
  4. Review the certificate setting and the required checks. Initially keep the temperature thresholds supplied by the BMC; different sensors have different acceptable temperatures.
  5. Save with Add Check. After the next run, open the result and check the model, sensor counts, Issues and notices about unavailable data.
  6. If needed, set up alerts for the check.

Settings

FieldTypeRequiredDefaultDescription
HostTextYesEmptyBMC hostname or IP address
BMC user / BMC passwordText / PasswordYesEmptyManagement-card credentials
HTTPS portIntegerNo443BMC port
Accept self-signed certificateToggleNoOnSkips server-certificate validation. Turn it off when the BMC uses a valid certificate trusted by the agent.
Check drivesToggleNoOnController, volume and drive health, failure prediction and SSD life remaining, where supplied by the BMC
Check voltagesToggleNoOnEvaluate voltages against the limits reported by the BMC
Check system event logToggleNoOnEvaluate warning and critical hardware-log entries, not login or audit logs
Event log window (hours)IntegerNo24Skip older dated entries; entries without a readable timestamp may still appear

Advanced options

Under Show advanced options, you can configure:

FieldTypeRequiredDefaultDescription
Use the thresholds of the management cardToggleNoOnUse each temperature sensor's own limits
Temperature warning (°C) / Temperature critical (°C)IntegersNo70 / 85Shared limits when BMC thresholds are disabled
Warn this many °C before the critical thresholdIntegerNo5Additional early warning near the critical limit; 0 disables it
SSD wear warning (% life left)IntegerNo15Warn when reported remaining life falls below this value
Ignore sensors (regular expression)TextNoEmptySkip matching temperature, fan and voltage sensors; matching ignores case

Results

The result uses the highest reported severity. Missing data does not prove healthy hardware:

  • No power-supply data: This query cannot detect a PSU failure. For redundant power supplies, add monitoring supported by the vendor, for example through IPMI.
  • No drives: Plain onboard SATA may expose no Redfish storage data. Add operating-system or controller monitoring. Virtual BMC media are not counted as physical drives.
  • No temperature or fan sensors: The check warns. Verify BMC support, firmware and access rights.
  • 0 RPM: This may mean a failed fan or an unpopulated header. If the BMC confirms a healthy sensor, or healthy overall status despite a degraded sensor status, the check reports a warning rather than critical. Inspect the server's header before excluding it. For two confirmed unpopulated headers, you could use ^(FRNT_FAN1|REAR_FAN1)$ as the sensor filter. Save and check the next run; a filter does not repair a fan.
  • Hardware events: A resolved problem may remain visible until it falls outside the time window. Large logs have evaluation limits; also inspect the complete BMC log when investigating a suspected fault.

Troubleshooting

If unreachable, check the BMC address, HTTPS port and firewall from the executing agent. For rejected credentials, check the account, password and permissions. No Redfish response requires checking whether the model supports Redfish; a reachable BMC web interface alone is not enough. An invalid regular expression prevents the check from running — correct or clear the sensor filter and run it again.