Amazon Q : Process exceeded memory threshold (support files.exclude, search.exclude vscode settings?)

## System details
- OS: Linux arm64 6.8.0-1026-aws
- Visual Studio Code extension host: 1.98.2 (Windows 11)
- Amazon Q: 1.53.0
- node: 20.18.2
- AWS Toolkit: 3.52.0

## Question

### Context

Hi All,
I'm working on a large Terragrunt & Terraform repository with over 50,000 files.
I'm running VSCode on Windows 11, connected to a remote AWS EC2 instance (r7g.medium with 8 GB RAM and 4 GB swap), where the Amazon Q extension is active.

### Issue

Amazon Q appears to scan the entire repository, which causes high memory usage on the EC2 instance. I've tried to reduce the scan scope using the settings.json file (both workspace and global) with files.exclude and search.exclude, but that doesn't seem to affect Amazon Q.

### Relevant Logs
```
2025-04-02` 11:03:03.281 [info] Command: (not started) [/home/myuser/.cache/aws/toolkits/language-servers/AmazonQ-Workspace/0.1.46/node-linux-arm64 /home/myuser/.cache/aws/toolkits/language-servers/AmazonQ-Workspace/0.1.46/qserver-linux-arm64/qserver/lspServer.js --nolazy --preserve-symlinks --stdio]
2025-04-02 11:03:03.336 [info] [Info - 11:03:03 AM] LSP server starts
2025-04-02 11:03:03.533 [info] [Info - 11:03:03 AM] Loaded model from /home/myuser/.cache/aws/toolkits/language-servers/AmazonQ-Workspace/0.1.46/qserver-linux-arm64/qserver
2025-04-02 11:03:03.688 [info] [Warn - 11:03:03 AM] Unknown tokenizer class "CodeSageTokenizer", attempting to construct from base class.
2025-04-02 11:03:03.688 [info] [Info - 11:03:03 AM] Using number of intra-op threads: 1
2025-04-02 11:03:03.910 [info] [Info - 11:03:03 AM] Embedding provider initialized.
2025-04-02 11:03:03.979 [warning] childProcess: Process 7444 exceeded memory threshold: 348217344
2025-04-02 11:03:03.979 [warning] childProcess: Process 7444 exceeded cpu threshold: 85.5
2025-04-02 11:03:13.975 [warning] childProcess: Process 7444 exceeded memory threshold: 348741632
2025-04-02 11:03:16.188 [info] auth: Connection is valid: 12345678-9abc-def-1234-567890abcd
2025-04-02 11:03:16.189 [info] auth: Updating connection state of 12345678-9abc-def-1234-567890abcd to valid
2025-04-02 11:03:23.992 [warning] childProcess: Process 7444 exceeded memory threshold: 348872704
2025-04-02 11:03:33.978 [warning] childProcess: Process 7444 exceeded memory threshold: 349003776
2025-04-02 11:03:43.985 [warning] childProcess: Process 7444 exceeded memory threshold: 349134848
2025-04-02 11:03:53.981 [warning] childProcess: Process 7444 exceeded memory threshold: 323571712
2025-04-02 11:04:03.984 [warning] childProcess: Process 7444 exceeded memory threshold: 323833856
2025-04-02 11:04:13.982 [warning] childProcess: Process 7444 exceeded memory threshold: 323964928
2025-04-02 11:04:23.983 [warning] childProcess: Process 7444 exceeded memory threshold: 324227072
2025-04-02 11:04:33.989 [warning] childProcess: Process 7444 exceeded memory threshold: 324358144
2025-04-02 11:04:43.987 [warning] childProcess: Process 7444 exceeded memory threshold: 324620288
2025-04-02 11:04:54.095 [warning] childProcess: Process 7444 exceeded memory threshold: 324882432
2025-04-02 11:05:03.996 [warning] childProcess: Process 7444 exceeded memory threshold: 325013504
2025-04-02 11:05:13.987 [warning] childProcess: Process 7444 exceeded memory threshold: 325275648
2025-04-02 11:05:23.988 [warning] childProcess: Process 7444 exceeded memory threshold: 324546560
2025-04-02 11:05:33.993 [warning] childProcess: Process 7444 exceeded memory threshold: 324546560
```

### My Question

Has anyone encountered this issue?

Is there a way to configure Amazon Q to limit which files or folders it indexes, to avoid overloading memory on the server?

Thanks in advance for any help!

David

Start by reproducing the scan with workspace and global settings.json using files.exclude and search.exclude, then inspect the Amazon Q qserver/lspServer.js entry point referenced in the logs. Done means determining whether those settings limit indexing and identifying a documented configuration or a scoped fix for repositories with many files.

Written by the indexing model from the issue text.