Connection Issues with Codeium Server Affecting Code Assistance

[wefor]

Description

Experiencing persistent connection issues with the Codeium server while using Cascade in Windsurf IDE. The connection keeps resetting, causing incomplete code modifications and increasing compilation errors.

Environment

• OS: macOS
• IDE: Windsurf
• Feature: Cascade (AI Code Assistant)
• Connection: From 192.168.2.202 to Codeium server (192.34.20.166:443)
• Time: 2025-01-21 20:00-20:50 GMT+8

Error Message

error executing cascade step: CORTEX_STEP_TYPE_CODE_ACTION: rpc error: code = Unknown desc = unavailable: read tcp 192.168.2.202:XXXXX->192.34.20.166:443: read: connection reset by peer

Steps to Reproduce

1. Open Windsurf IDE
2. Use Cascade to assist with code modifications
3. Attempt to make multiple code changes
4. Observe frequent connection drops during code editing operations

Current Behavior

• Connection to Codeium server frequently resets
• Code edits are terminated mid-way
• Cascade cannot complete its assistance tasks
• Results in incomplete code changes
• Leads to increasing compilation errors
• Multiple occurrences within an hour

Expected Behavior

• Stable connection to Codeium server
• Complete code modifications without interruption
• Consistent assistance from Cascade

Impact

The connection issues are severely affecting the ability to use Cascade effectively:

1. Cannot complete code modifications
2. Increases development time due to incomplete changes
3. Creates additional work to fix partial modifications
4. Reduces the reliability of the AI assistance feature

Additional Context

• The issue occurs consistently during code editing operations
• Each disconnection requires restarting the modification process
• The error pattern is consistent, always showing "connection reset by peer"

Logs

Step Id: XXX Encountered error in step execution: error executing cascade step: CORTEX_STEP_TYPE_RETRIEVE_MEMORY: read tcp 192.168.2.202:XXXXX -> 192.34.20.166:443: read: connection reset by peer

Questions

1. Is this a known issue?
2. Are there any available workarounds?
3. Is there a way to make the connection more resilient?
4. Can the system be modified to resume interrupted operations?

Possible Related Issues

• Search for similar connection reset issues
• Check if other users have reported similar problems

---

Please let me know if you need any additional information to investigate this issue.