Servers Forensic Technical Intake: Workstation Server Systems Service Protocol: Dual-Socket Architecture & High-Wattage Modular Swap 1. Primary Fault Classification Fault Group (A-Z) — Select Group — Compute (Dual CPU / ECC Memory) Graphics (Multi-GPU / GPGPU / NVLink) Logic Board (Backplane / PCIe Bridges) Management (IPMI / iDRAC / ILO) Power (Redundant PSU / PDB Module) Storage (Hardware RAID / HBA / NVMe) Thermal (High-Static Pressure Fans) Specific Fault — Select Group First — Describe the forensic symptoms: 2. Server Architecture & Power Redundancy Forensic Power Analysis: Workstation servers often utilize **Redundant PSUs**. If the system powers on but warns of “Lost Redundancy,” the modular **Power Distribution Board (PDB)** or a specific PSU module has failed. We must verify if the system can sustain peak AI compute loads without tripping the OCP (Over-Current Protection). PSU Configuration Single High-Wattage Dual Redundant (1+1) Quad Redundant (2+2 / 3+1) Describe any IPMI/iDRAC/ILO error codes or LCD panel messages: 3. Multi-Socket & ECC Memory Integrity AI Technical Insight: In dual-socket workstation servers, a “Dead System” can often be traced to a single failing **CPU Interconnect**. If the system boots with 128GB but only sees 64GB, a modular **ECC RAM stick** or the **CPU Integrated Memory Controller (IMC)** is the failing component. Total Installed ECC RAM (e.g., 256GB DDR4 ECC): Memory/Compute Symptoms (A-Z) Full Memory Capacity Recognized Specific DIMM Slot Disabled CPU Socket 2 Not Initializing Uncorrectable ECC Error (HALT) 4. Storage Backplane & RAID Integrity Forensic Note: Workstation servers use a **Backplane module** to bridge the drives to the Logic Board. If multiple drives drop offline simultaneously, the **SAS Expander** or the **RAID Controller module** is the failure point. Shipping these units with drives installed can result in backplane connector shearing. Storage Configuration Direct Pass-through (HBA) Hardware RAID (With Cache Battery) Software RAID (ZFS/TrueNAS) Describe any ‘Drive Missing’ or ‘Predictive Failure’ status lights: 5. Previous Repair & Modification History Why we need this: Servers are often upgraded with third-party PCIe NICs or NVMe adapters. If the **PCIe Lane bifurcation** was configured incorrectly in the BIOS, it can cause the system to fail “Forensic Power Analysis” due to resource conflicts. Has this Server ever been repaired or modified? No – Factory Original Yes – Repaired/Upgraded Previously Describe previous work (e.g., ‘Added 10GbE SFP+ card’, ‘Repasted CPUs’): 6. Mandatory Packaging Compliance IMPORTANT: Workstation Servers are extremely heavy (40lb+). The responsibility for secure packing remains entirely with the customer. You MUST use a heavy-duty, double-walled box with a minimum of 4 inches of high-density foam on all sides. REMOVE any modular GPUs or mechanical HDDs and ship them in separate anti-static packaging to prevent internal structural collapse. Confirm internal components are braced and specify box strength: