Bad performance feedback loop with slow PSScriptAnalyzer rules

Beginnen Sie mit dem Script-Analysis-Scheduling-Pfad rund um Invoke-ScriptAnalyzer und den im Issue beschriebenen pssaRunspacePool mit Einzelkonkurrenz. Reproduzieren Sie die vier verzögerten Bearbeitungen mit einer benutzerdefinierten Regel, die etwa fünf Sekunden benötigt, und verfolgen Sie anschließend, wie ausstehende Scans in die Warteschlange gestellt werden. Als abgeschlossen gilt die Aufgabe, wenn neuere Dokumentänderungen Scans vorziehen, die noch nicht gestartet wurden, und die Ergebnisse beim aktuellen Dokument konvergieren, ohne jede Zwischenversion zurückzugeben.

Prerequisites

• I have written a descriptive issue title.
• I have searched all open and closed issues to ensure it has not already been reported.
• I have read the troubleshooting guide.
• I am sure this issue is with PowerShell Editor Services itself and does not reproduce in a standalone PowerShell instance, and is not an issue with my editor.
• I have verified that I am using the latest version of PowerShell Editor Services.
• If this is a security issue, I have read the security issue reporting guidance.

Summary

I'm running into a script analysis performance issue in my environment. We have a decent number of custom PSScriptAnalyzer rules that can take more than several seconds to complete with larger scripts.

As you're typing in a PowerShell document, the responsiveness of script analysis deteriorates to the point were you have to wait many times more than the usual run duration of the rules to see most up-to-date analysis results.

This is happening because every time you stop typing for >750ms PSES fires off another multi-second Invoke-ScriptAnalyzer run into the pssaRunspacePool (which has a concurrency of 1). So with a lot of furious typing, script analysis service can end up with tens of seconds of work just to catch-up to the document contents that are on the screen at the moment.

I've come up with a fix that boils down to holding off Invoke()'ing a new scan until the currently running one completes. That gives a chance for any more incoming document changes to pre-empt ones that have not started scanning yet. As a result, the scanning results delay is more or less stable and not worse than the runtime of all of the rules.

I need to do more testing of the fix, but was wondering if anyone else has run into this? Maybe there would be a better way to fix it? Planning to submit a PR for this at some point.

PowerShell Version

ame                           Value
----                           -----
PSVersion                      7.4.11
PSEdition                      Core
GitCommitId                    7.4.11
OS                             Microsoft Windows 10.0.22631
Platform                       Win32NT
PSCompatibleVersions           {1.0, 2.0, 3.0, 4.0…}
PSRemotingProtocolVersion      2.3
SerializationVersion           1.1.0.1
WSManStackVersion              3.0

Editor Version

VSCode
------
Version: 1.104.3 (system setup)
Commit: 385651c938df8a906869babee516bffd0ddb9829
Date: 2025-10-02T12:30:51.747Z
Electron: 37.3.1
ElectronBuildId: 12404162
Chromium: 138.0.7204.235
Node.js: 22.18.0
V8: 13.8.258.31-electron.0
OS: Windows_NT x64 10.0.22631

PowerShell Editor Services Version

Major  Minor  Build  Revision
-----  -----  -----  --------
4      3      0      0

Steps to Reproduce

1. Create a custom PSScriptAnalyzer rule that takes 5s to run (easier to reproduce the issue with longer delays).
2. Type, wait 1 second, type, wait 1 second, type, wait 1 second, type.
3. Observe that after stopping typing script analysis will continue running for the next ~15s and will return all of the intermediate results as you were typing. As an interactive user I only care about analysing the document that is on the screen now.

Visuals

No response

Logs

No response