Tier 2 · SOP-007 · Module 07 of 09

Windows Performance Diagnostics

Diagnose slow, unresponsive, and high-resource machines using Task Manager, Resource Monitor, and Event Viewer. Read the data, find the bottleneck, fix it.

Tier 2 — IT Support SOP-007 Intermediate Diagnostics
📊

Phase 1 — Learn IT

Understand the four performance bottleneck categories and the tools that measure each one.

The Four Performance Bottlenecks

Every slow computer complaint falls into one of four categories: CPU, Memory (RAM), Disk (Storage), or Network. Your job is to identify which resource is the bottleneck — then either fix it or escalate for a hardware upgrade.

Randomly disabling things or doing disk cleanups without checking metrics first is guesswork. Read the numbers, then act.

ResourceSymptomTool to CheckCritical Threshold
CPUSystem unresponsive, apps freeze during tasksTask Manager → Performance → CPUSustained >85% usage
RAMConstant freezing, very slow app switchingTask Manager → Performance → Memory>90% in use; lots of disk paging
DiskExtremely slow boot, apps take forever to openTask Manager → Performance → Disk100% active time (disk bottleneck)
NetworkSlow web browsing, file transfers, video callsTask Manager → Performance → Network / Resource MonitorConsistently near bandwidth cap

Your Performance Toolkit

ToolHow to OpenWhat It Shows
Task ManagerCtrl+Shift+Esc or right-click taskbarPer-process CPU/RAM/Disk/Network usage; Startup tab; Performance graphs
Resource MonitorTask Manager → Performance → Open Resource MonitorDetailed per-process and per-disk-drive breakdown; network connections by process
Performance MonitorRun → perfmonConfigurable real-time performance counters and data collector sets
Event ViewerRun → eventvwrSystem/Application error logs; critical events and crash analysis
Reliability MonitorControl Panel → Security → View Reliability HistoryTimeline of software installs, crashes, and errors; identifies when problems started
msinfo32Run → msinfo32Complete hardware inventory: CPU, RAM amount/speed, installed software, drivers
💡

Reliability Monitor Is Underused

Most techs skip Reliability Monitor, but it's incredibly useful for timeline diagnosis. It shows a calendar view of crashes, errors, warnings, and software installs. When a user says "it started being slow last Tuesday," open Reliability Monitor to see exactly what happened on that date — often a bad update or new software install will be right there.

What's Actually Slowing the Machine

CauseBottleneck TypeFix
Too many startup programsCPU & RAMTask Manager → Startup → disable unnecessary items
Insufficient RAM for workloadRAM → Disk (paging)Close apps; recommend RAM upgrade if chronic
Spinning HDD (vs. SSD)DiskSSD upgrade dramatically improves boot/load times
HDD near capacity (<10% free)DiskDisk Cleanup + offload files; paging file needs space
Malware / cryptominerCPU / DiskIdentify process in Task Manager; run malware scan (SOP-006)
Windows Update running in backgroundCPU / Disk / NetworkCheck Windows Update status; let it complete or schedule
Corrupt or fragmented disk (HDD)DiskRun chkdsk /f; optimize drive (Defragment & Optimize)
Thermal throttling (overheating)CPUCheck temps with HWMonitor; clean dust from vents/fans

Key Terms

Paging File
A file on disk (pagefile.sys) Windows uses as overflow memory when RAM is full. Heavy paging causes severe slowdowns because disk is 100x slower than RAM.
Thermal Throttling
CPU automatically reduces speed to prevent overheating. Causes slowdowns that look like software issues. Clean dust from vents as first physical check.
Disk I/O
Input/Output operations on storage. When disk I/O is at 100%, nothing can read or write — system appears frozen waiting for disk access.
Event ID
Numeric codes in Event Viewer that identify specific events. Event ID 41 = unexpected shutdown/crash. Event ID 1001 = Windows Error Reporting (crash dump).
Handle / Thread
Handles are references to system resources (files, registry keys) held by a process. A process with thousands of handles may have a resource leak — look for it in Resource Monitor.
📋

Phase 2 — Do IT

Execute SOP-007. Measure first, fix second. Never skip the verification step.

SOP-007 · WINDOWS PERFORMANCE DIAGNOSTICS · REV 1.0

Slow / Unresponsive Windows System Diagnosis

Use this procedure for any ticket where a user reports the system is "slow," "freezing," or "taking forever." Always collect data first before making any changes.

01

Open Task Manager and Read the Performance Tab

Press Ctrl+Shift+Esc and click the Performance tab. Look at all four: CPU %, Memory used/available, Disk active time %, and Network throughput. Identify which resource is at or near 100%. This tells you your bottleneck category before you touch anything.

02

Identify the Top Consuming Processes

In Task Manager's Processes tab, click the column header for the bottleneck resource (CPU, Memory, or Disk) to sort by usage. Identify the top processes consuming that resource. Right-click a suspicious process → Open file location to see if it's a legitimate Windows process or something unknown.

03

Check Startup Programs

Click the Startup tab in Task Manager. Sort by "Startup impact." Disable any unnecessary programs that launch at boot — especially anything with "High" impact that the user doesn't recognize. This alone often resolves boot slowness significantly. Requires a reboot to take effect.

04

Check Available Disk Space

Open File Explorer → This PC. Check each drive's free space. If the system drive (C:) has less than 10% free, disk operations suffer and the paging file has limited room. Run Disk Cleanup (search for it) and clean Windows Update files, temporary files, and Recycle Bin. If still critically low, escalate for storage expansion.

05

Check Event Viewer for Critical Errors

Open Event Viewer → Windows Logs → System. Filter by Error and Critical. Look for recurring errors or errors that coincide with the slowness. Specifically look for Event ID 41 (unexpected shutdown) and Disk errors from the source "Disk" — these can indicate impending drive failure.

06

Run Disk Health Check

Open Command Prompt as Administrator and run chkdsk C: /f /r. If the drive is in use, schedule it for next reboot when prompted. For SSDs or HDDs, also check SMART data using a tool like CrystalDiskInfo (free) — look for "Caution" or "Bad" status, which indicates drive failure is imminent.

07

Check Windows Update and Ensure System Is Patched

Sometimes Windows Update runs silently in the background consuming resources. Open Settings → Windows Update and check if an update is actively downloading or installing. Let it finish, then reboot. After updating, run sfc /scannow in an elevated Command Prompt to repair any corrupted system files that may be causing instability.

Escalate / Recommend Hardware When

  • RAM consistently at >90% during normal workload — user needs a RAM upgrade
  • System drive is a spinning HDD and all software fixes have been tried — recommend SSD upgrade
  • SMART data shows imminent drive failure — replace drive immediately, escalate urgently
  • Event Viewer shows recurring CPU or memory hardware errors (not software-related)
  • CPU is sustained at 100% with no user processes visible — possible hardware fault or malware (SOP-006)

Performance Cheat Sheet

BottleneckFirst CheckCommon Fix
CPU 100%Task Manager → Processes sorted by CPUEnd rogue process; check for malware; check cooling
RAM 90%+Task Manager → Memory; paging activityDisable startup items; close apps; recommend RAM upgrade
Disk 100%Task Manager → Disk; identify top I/O processFree disk space; chkdsk; SSD upgrade for HDD
Slow bootStartup tab in Task ManagerDisable high-impact startup items
Slow after updateReliability Monitor; Event ViewerRoll back update or run sfc /scannow
🎯

Phase 3 — Apply IT

Diagnose performance tickets from data, not guesswork. Show your methodology clearly.

Practice Tickets

SCENARIO 01

The Slow Bootup

"My computer takes about 10 minutes to fully boot up. Once it's on it's fine, but startup is unbearable. It's been getting worse over the past few months."
Your Tasks
  • 1.What is the first diagnostic tab you open and what are you looking for?
  • 2.You see 12 startup items — 6 have "High" impact including four third-party apps. What do you do?
  • 3.After disabling startup items and rebooting, it still takes 5 minutes. Task Manager shows Disk at 100% during boot. What is the likely hardware cause and recommendation?
  • 4.How do you check if the drive is failing vs. just being slow because it's a spinning HDD?
SCENARIO 02

The Freezing Workstation

"My computer freezes for 10–15 seconds multiple times a day. When it unfreezes, everything goes back to normal. It doesn't crash, just locks up temporarily."
Your Tasks
  • 1.Task Manager shows Memory at 95% during the freeze. What is happening physically and what are the two options?
  • 2.The machine has 8GB RAM. What question do you ask about the user's workload to determine if an upgrade is warranted?
  • 3.You check Event Viewer and find several Disk errors from the past week. What do you investigate next?
  • 4.Write a ticket note explaining what you found and your recommendation.
SCENARIO 03

The Mystery CPU Spike

"My fan is running at full speed constantly and the computer is very slow even when I'm not doing anything. I haven't installed anything recently."
Your Tasks
  • 1.High fan speed + high idle CPU — what are the two most likely causes?
  • 2.Task Manager shows a process called "Service Host: SysMain" at 40% CPU. Is this malware or a Windows process? What does it do?
  • 3.You check and Windows Update is in the middle of a large feature update. How does this change your approach?
  • 4.After the update completes and CPU returns to normal, what do you check before closing the ticket?

Document Your Resolution

FieldYour Entry
Bottleneck Found
Metrics Observed
Root Cause
Steps Taken
Resolution / Recommendation

Questions You'll Face

Interview Question 01

"A user says their computer is running slowly. Walk me through how you diagnose it."

Strong Answer Framework

I start by opening Task Manager and checking the Performance tab — CPU, Memory, Disk, and Network — to identify which resource is the bottleneck. Once I know the bottleneck, I sort the Processes tab by that metric to find the top consumer. If it's a startup issue, I check the Startup tab for high-impact programs. If disk is at 100%, I check free space and run a disk health check. If RAM is maxed, I look at whether the workload genuinely exceeds the machine's capacity. I also check Reliability Monitor to see if any events coincide with when the slowness started.

Interview Question 02

"What does it mean when the disk is at 100% utilization and how do you fix it?"

Strong Answer Framework

100% disk utilization means the drive is at capacity for read/write operations — everything waiting to read or write is queued, which causes the system to appear frozen. Common causes are: the paging file growing because RAM is full, Windows Update downloading in the background, a failing drive with bad sectors, a malware process, or just the inherent limitation of a spinning HDD which is mechanically slow. I check the Disk column in Task Manager Processes to find what's generating the I/O. If it's Windows Update, I wait. If it's the system paging heavily, RAM may need an upgrade. If the drive shows high response times (>50ms average) or SMART errors, a hardware replacement is needed.

Interview Question 03

"What is the difference between Task Manager and Resource Monitor, and when do you use each?"

Strong Answer Framework

Task Manager gives you a high-level real-time view — which processes are using the most CPU, RAM, disk, and network, plus startup items and performance graphs. I use it as the first tool to identify the bottleneck category and the suspect process. Resource Monitor is deeper — launched from Task Manager's Performance tab, it shows per-process breakdown by specific disk file access, per-connection network activity, and memory commit details. I use Resource Monitor when Task Manager shows an issue but I need to know exactly which files a process is accessing, or which process is holding a specific network connection open. It's the difference between "something is using the disk" and "here is the exact file and operation."

Completion Checklist

I can identify the four bottleneck types (CPU/RAM/Disk/Network) using Task Manager
I know when to use Task Manager vs. Resource Monitor vs. Event Viewer
I can disable startup items and explain why that improves boot time
I understand paging and why 100% disk usage is often a RAM problem
I know how to check SMART data for drive health and what the results mean
I completed all three scenarios and documented at least one with metrics

Continue Your Training

Module 07 · SOP-007
Finished all three phases? Mark this module complete.