Operator Monitor And CPU Affinity Tools Reference

Operator Monitor And CPU Affinity Tools Reference

Verified 2026-06-21 against master 0139a346 (Arma 2 OA 1.64); paths are repo-root relative for Tools/ and docs/testing/ sources.

This page covers two small operator-side scripts: Tools/Monitor/Get-WindowedRpt.ps1, a reusable current-window RPT reader, and Tools/Ops/Set-WaspCpuAffinity.ps1, a dry-run-first CPU-affinity helper for the dedicated server and headless clients (Tools/Monitor/Get-WindowedRpt.ps1:1-15; Tools/Ops/Set-WaspCpuAffinity.ps1:1-17).

Tool Inventory

Tool	Entry point	Primary use	Source evidence
Tools/Monitor/Get-WindowedRpt.ps1	PowerShell function Get-WindowedRpt	Read only the RPT lines since the most recent mission or boot marker, then optionally regex-filter and tail the result.	Tools/Monitor/Get-WindowedRpt.ps1:1-7, Tools/Monitor/Get-WindowedRpt.ps1:15-24, Tools/Monitor/Get-WindowedRpt.ps1:47-58
Tools/Ops/Set-WaspCpuAffinity.ps1	Script parameters -ServerMask, -HcMasks, -Apply	Report or set processor-affinity masks for arma2oaserver.exe and headless-client ArmA2OA.exe processes.	Tools/Ops/Set-WaspCpuAffinity.ps1:13-17, Tools/Ops/Set-WaspCpuAffinity.ps1:39-44, Tools/Ops/Set-WaspCpuAffinity.ps1:71-87
docs/testing/hc-scaling-test.md	Test methodology document	Explains why the affinity helper exists in the 0/1/2-HC scaling test, and records that the script is prepped but not applied.	docs/testing/hc-scaling-test.md:1-8, docs/testing/hc-scaling-test.md:55-75, docs/testing/hc-scaling-test.md:77-81

Windowed RPT Reader

Get-WindowedRpt exists because the script comments treat A2 OA RPT files as append-only during a server lifetime; whole-file greps become slower and can match stale errors from earlier missions (Tools/Monitor/Get-WindowedRpt.ps1:1-7). Its intended usage is dot-sourcing the script and calling Get-WindowedRpt for a current mission window, an error regex, or a custom boot marker such as Dedicated host created (Tools/Monitor/Get-WindowedRpt.ps1:9-13).

Behavior	Details	Source evidence
Required input	-RptPath is mandatory.	Tools/Monitor/Get-WindowedRpt.ps1:17-18
Default window	-WindowMarker defaults to MISSINIT, described as the mission init marker stamped at every mission start.	Tools/Monitor/Get-WindowedRpt.ps1:19-20
Optional filtering	-Pattern is an optional regex applied to lines inside the selected window.	Tools/Monitor/Get-WindowedRpt.ps1:21-22, Tools/Monitor/Get-WindowedRpt.ps1:54
Optional tailing	-Tail returns at most the requested number of lines from the window end; 0 means all lines.	Tools/Monitor/Get-WindowedRpt.ps1:23-24, Tools/Monitor/Get-WindowedRpt.ps1:55-57
Missing file behavior	If the RPT path is absent, the function warns and returns an empty array.	Tools/Monitor/Get-WindowedRpt.ps1:27-30
Non-locking read	The script opens the RPT with FileAccess.Read and FileShare.ReadWrite, then reads it through a StreamReader.	Tools/Monitor/Get-WindowedRpt.ps1:32-43
Window selection	The function scans backward for the last marker match and returns lines from that marker to EOF; if no later marker is found, it returns the full file.	Tools/Monitor/Get-WindowedRpt.ps1:45-52
Return value	The function returns the selected window after optional regex and tail steps.	Tools/Monitor/Get-WindowedRpt.ps1:54-58

CPU Affinity Helper

Set-WaspCpuAffinity.ps1 is prepared for the 0/1/2-HC scaling test, where the test plan compares zero, one, and two headless clients across day-long runs while keeping other lobby parameters constant (docs/testing/hc-scaling-test.md:1-8, docs/testing/hc-scaling-test.md:32-44). The CPU-affinity section says the main server has 6 P-cores plus E-cores, and the mitigation is to pin the server and headless clients to dedicated P-core sets so the Arma processes do not fight over shared cache (docs/testing/hc-scaling-test.md:55-60).

Behavior	Details	Source evidence
Dry-run default	The synopsis says the script prints what it would do unless -Apply is passed.	Tools/Ops/Set-WaspCpuAffinity.ps1:17-18, Tools/Ops/Set-WaspCpuAffinity.ps1:32-37
Parameters	-ServerMask defaults to 0, -HcMasks defaults to an empty array, and -Apply controls whether masks are set.	Tools/Ops/Set-WaspCpuAffinity.ps1:39-44
Server target	The server process target is arma2oaserver.exe; if no server mask is provided, a found server is left untouched.	Tools/Ops/Set-WaspCpuAffinity.ps1:13-15, Tools/Ops/Set-WaspCpuAffinity.ps1:71-76
Headless-client target	Headless clients are ArmA2OA.exe processes whose command line contains -client, sorted by creation date.	Tools/Ops/Set-WaspCpuAffinity.ps1:13-16, Tools/Ops/Set-WaspCpuAffinity.ps1:46-51
HC mask order	The script applies one -HcMasks entry per HC in connection order; HCs beyond the supplied masks are left untouched.	Tools/Ops/Set-WaspCpuAffinity.ps1:28-30, Tools/Ops/Set-WaspCpuAffinity.ps1:78-87
Actual write	When -Apply is present, Set-Affinity writes ProcessorAffinity on the process and prints the old and new mask.	Tools/Ops/Set-WaspCpuAffinity.ps1:53-64
Error handling	Affinity write failures are caught and emitted as PowerShell warnings.	Tools/Ops/Set-WaspCpuAffinity.ps1:64-66
Mask semantics	Masks are hardware-specific bitmasks over logical processors; bit 0 is CPU0, and the script comment gives a 6-P-core hyperthreading example.	Tools/Ops/Set-WaspCpuAffinity.ps1:19-23
Test-plan status	The test document marks the affinity script as prepped and not applied.	docs/testing/hc-scaling-test.md:77-81

Operator Boundaries

Boundary	Practical reading	Source evidence
RPT helper scope	Get-WindowedRpt reads an existing RPT and returns selected lines; it does not parse PerformanceAudit tables, launch Arma, or restart a server.	Tools/Monitor/Get-WindowedRpt.ps1:15-24, Tools/Monitor/Get-WindowedRpt.ps1:32-58
Affinity helper scope	Set-WaspCpuAffinity.ps1 discovers existing Arma processes and only changes ProcessorAffinity when -Apply is present.	Tools/Ops/Set-WaspCpuAffinity.ps1:46-66, Tools/Ops/Set-WaspCpuAffinity.ps1:71-87
Live-host caution	The HC test note says masks are hardware-specific, must be confirmed on the target host, and should not be applied on the live host without a maintenance window and explicit go-ahead.	docs/testing/hc-scaling-test.md:72-75
Measurement boundary	The HC-scaling plan measures client FPS, server FPS, HC FPS/load, and group-count signals; the affinity helper is one mitigation item inside that test, not the test result.	docs/testing/hc-scaling-test.md:10-18, docs/testing/hc-scaling-test.md:55-63

Quick Commands

Task	Command shape	Source evidence
Load the RPT helper	. C:\WASP\monitor\Get-WindowedRpt.ps1	Tools/Monitor/Get-WindowedRpt.ps1:9-10
Read the current mission window	$lines = Get-WindowedRpt -RptPath $rpt	Tools/Monitor/Get-WindowedRpt.ps1:11
Read matching error lines	$errs = Get-WindowedRpt -RptPath $rpt -Pattern 'Error|ERROR'	Tools/Monitor/Get-WindowedRpt.ps1:12
Read a per-boot window	$boot = Get-WindowedRpt -RptPath $rpt -WindowMarker 'Dedicated host created'	Tools/Monitor/Get-WindowedRpt.ps1:13
Dry-run CPU affinity	.\Set-WaspCpuAffinity.ps1 -ServerMask 0x0FF -HcMasks 0x300,0xC00	Tools/Ops/Set-WaspCpuAffinity.ps1:35-37, docs/testing/hc-scaling-test.md:65-69
Apply CPU affinity	.\Set-WaspCpuAffinity.ps1 -ServerMask 0x0FF -HcMasks 0x300,0xC00 -Apply	Tools/Ops/Set-WaspCpuAffinity.ps1:35-37, docs/testing/hc-scaling-test.md:65-69

Continue Reading

• Tools and build workflow
• Server ops runbook
• PerformanceAuditAnalyzer
• Headless client scaling and topology
• Testing, debugging and release workflow