What is field replaceable unit (FRU)?
In the high-stakes world of data center maintenance, downtime is measured in dollars per second. When a component fails, you don't want to send the entire server back to the manufacturer or wait for a specialized engineer to desolder a chip. You want to swap the part and get back to business.
This is where the Field Replaceable Unit (FRU) comes in.
A Field Replaceable Unit is a modular component of a system that is designed to be quickly and easily replaced by a technician (or even an end-user) at the location where the equipment is installed.
In Oracle and other enterprise hardware, a FRU is not just a piece of plastic and metal; it is a digitally tracked asset. Every FRU contains a tiny chip called an ID PROM (Programmable Read-Only Memory) that stores:
The Part Number
The Serial Number
The Manufacturer and Date
The Hardware Revision level
Not all FRUs are created equal. They are generally categorized by how much "interruption" they cause during replacement:
These are the gold standard for high availability. You can pull these out while the server is running without any impact on the OS.
Examples: Hard drives, Power Supply Units (PSUs), and some Fan modules.
The Benefit: Zero downtime. The system detects the removal and the new insertion automatically.
These require the system to be powered down before replacement to prevent electrical damage or system crashes.
Examples: CPUs, Memory (DIMMs), and Motherboards.
The Benefit: Modular repair is still faster and cheaper than replacing the whole server, even if a brief maintenance window is required.
When a server boots up, the Service Processor (SP) performs an "inventory walk." It talks to the ID PROM on every single FRU to ensure everything is compatible.
If a component fails, the system uses this data to help the technician:
The Amber LED: The SP sends a signal to light a physical LED on the specific FRU that is failing.
The Logic Trace: In the logs, the system won't just say "Memory Error"; it will say "Fault in FRU: /SYS/MB/DIMM0". This tells the technician exactly which slot to target.
A typical enterprise FRU replacement follows this "blind-simple" workflow:
Step 1: Identification: The Automated Diagnostic Framework (ADF) identifies a fault and lights the "Service Required" LED.
Step 2: Verification: The technician runs a command (like fmadm faulty) to confirm the Part Number.
Step 3: Swap: The technician slides the old FRU out and clicks the new one in.
Step 4: Auto-Update: The Service Processor detects the new Serial Number, verifies the firmware is compatible, and clears the fault automatically.
The FRU is the building block of serviceability. By making a server modular, manufacturers ensure that hardware failures are a minor logistics task rather than a catastrophic engineering project. If your server doesn't have a clear FRU strategy, you aren't running an enterprise data center—you're running a risk.