5.3.1. Storage Failure Symptoms and Causes
💡 First Principle: Storage failure symptoms often cascade — a single failed drive causes RAID degradation, which causes performance problems, which appear as application slowness. Tracing the cascade backward to the physical root cause requires understanding the dependency chain from RAID through volume through file system through application.
Common Storage Problems and Their Causes
| Symptom | Most Likely Cause | Urgency |
|---|---|---|
| OS not found at boot | Boot sector corruption, OS partition failed, boot order wrong | Critical |
| Drive not available | Drive failed, cable disconnected, HBA failure | Critical |
| Multiple drive failure | RAID 5 losing second drive during rebuild, backplane failure | Critical — data loss imminent |
| Slow file access | RAID rebuilding in degraded mode, drive near failure, cache failure | High |
| Slow I/O performance | IOPS saturation, RAID write penalty, cache turned off | High |
| Boot errors | Corrupt MBR/GPT, failed boot drive, missing boot files | High |
| Sector/block errors | Drive developing bad sectors, filesystem corruption | Medium — escalating |
| Cache battery failure | RAID controller cache running in write-through (slower) mode | Medium |
| Partition errors | Improper partition table, filesystem not mounted | Medium |
| Data corruption | Bad sectors, incomplete writes during crash, bit rot | Medium-High |
| Unsuccessful backup | Drive full, permissions, open file lock, network issue | Medium |
| Unable to mount device | Filesystem corruption, wrong filesystem type, missing drivers | Medium |
Root Cause Analysis for Storage Failures
| Root Cause | What It Affects | How to Identify |
|---|---|---|
| Disk space exhaustion | Writes fail; services crash | df -h (Linux), Disk Management (Windows) |
| Misconfigured RAID | Array won't initialize or rebuilds incorrectly | RAID controller management console |
| Media failure | Specific files unreadable; SMART errors | SMART monitoring, drive diagnostics |
| Controller failure | All drives on controller inaccessible | Replace controller; verify with known-good controller |
| HBA failure (SAN) | SAN LUNs disappear | Check HBA status, port lights, fabric connectivity |
| Loose connectors | Intermittent drive visibility | Physical inspection, re-seat cables |
| Corrupt boot sector | OS won't start; "OS not found" | Boot repair tools (bootrec, grub-install) |
| Bad sectors | Specific file read failures; growing SMART error count | SMART data, chkdsk /r, badblocks |
| Array rebuild in progress | Degraded performance; RAID rebuilding | RAID management console |
| Improper RAID config | Array not created correctly | Verify RAID level, stripe size, drive membership |
| Mismatched drives | RAID won't accept drive; performance issues | Match drive model/capacity within array |
| Cache battery failure | Controller falls back to write-through | Replace RAID controller battery; performance degrades |
⚠️ Exam Trap: A RAID array "rebuilding" is not a healthy state — it's a degraded state that is recovering. During rebuild, the array is vulnerable to total data loss if another drive fails. The exam may present a rebuilding array as the root cause of slow performance while the real test is recognizing the data loss risk.
Reflection Question: A server with a RAID 5 array of four drives shows the array in "degraded" status in the RAID management console, with one drive shown as "failed." A new drive is hot-swapped in and the rebuild begins. Three hours into the rebuild, a second drive shows errors. What is the current status of the data, and what should have been done proactively before starting the rebuild?
