chore: add error log on model prediction failure

[yamoyamoto]

User description

Add error message to logging for model prediction failures in retry_with_fallback_models, capturing the exception message.

I’m currently encountering PR-Agent issue #2042 and need richer logs to investigate. No functional changes; logging-only.

---

PR Type

Enhancement

---

Description

• Add exception details to model prediction failure logs

• Improve error handling in retry_with_fallback_models function

• Capture specific exception messages for debugging purposes

---

Diagram Walkthrough

flowchart LR
  A["Model Prediction"] -- "Exception occurs" --> B["Log Warning"]
  B -- "Enhanced with" --> C["Exception Message"]

Loading

File Walkthrough

	Relevant files
Enhancement	pr_processing.py Enhanced exception logging in retry function                          pr_agent/algo/pr_processing.py Enhanced exception handling to capture exception instance Added exception message to warning log output Improved debugging information for model prediction failures +2/-2

---

[qodo-merge-for-open-source]

PR Compliance Guide 🔍

(Compliance updated until commit a194ca6)

Below is a summary of compliance checks for this PR:

Security Compliance
🟢	No security concerns identified No security vulnerabilities detected by AI analysis. Human verification advised for critical code.
Ticket Compliance
🟡	🎫 #2042 🟢 Add detailed error logging for model failures, including the exception message. Improve visibility into which model failed during retries. Preserve or improve final aggregated failure message when all fallback models fail. 🔴 ⚪ Investigate large PR processing with different models (requires runtime validation). Verify ARM64 host with AMD64 container compatibility (requires environment/runtime validation).
Codebase Duplication Compliance
⚪	Codebase context is not defined Follow the guide to enable codebase context checks.
Custom Compliance
🟢	Consistent Naming Conventions Objective: All new variables, functions, and classes must follow the project's established naming standards Status: Passed
	No Dead or Commented-Out Code Objective: Keep the codebase clean by ensuring all submitted code is active and necessary Status: Passed
	Robust Error Handling Objective: Ensure potential errors and edge cases are anticipated and handled gracefully throughout the code Status: Passed
	Single Responsibility for Functions Objective: Each function should have a single, well-defined responsibility Status: Passed
	When relevant, utilize early return Objective: In a code snippet containing multiple logic conditions (such as 'if-else'), prefer an early return on edge cases than deep nesting Status: Passed

Compliance status legend

🟢 - Fully Compliant
🟡 - Partial Compliant
🔴 - Not Compliant
⚪ - Requires Further Human Verification
🏷️ - Compliance label

---

Previous compliance checks

Compliance check up to commit be93651

Security Compliance
🟢	No security concerns identified No security vulnerabilities detected by AI analysis. Human verification advised for critical code.
Ticket Compliance
🟡	🎫 #2042 🟢 Add detailed error logging when model prediction fails, including the exception message. Improve visibility into which model/deployment failed during retries. Enable investigation of large PR processing issues by enriching logs without changing functionality. 🔴 Consider environment/compatibility factors in investigation (ARM64 host vs AMD64 container) – may require diagnostic logging. ⚪ None
Codebase Duplication Compliance
⚪	Codebase context is not defined Follow the guide to enable codebase context checks.
Custom Compliance
🟢	Consistent Naming Conventions Objective: All new variables, functions, and classes must follow the project's established naming standards Status: Passed
	No Dead or Commented-Out Code Objective: Keep the codebase clean by ensuring all submitted code is active and necessary Status: Passed
	Robust Error Handling Objective: Ensure potential errors and edge cases are anticipated and handled gracefully throughout the code Status: Passed
	Single Responsibility for Functions Objective: Each function should have a single, well-defined responsibility Status: Passed
	When relevant, utilize early return Objective: In a code snippet containing multiple logic conditions (such as 'if-else'), prefer an early return on edge cases than deep nesting Status: Passed

[qodo-merge-for-open-source]

PR Code Suggestions ✨

Latest suggestions up to a194ca6

Category	Suggestion	Impact
General	Log full, serializable error context Enhance error logging by serializing the exception object, adding the stack traceback, and including more context such as deployment ID and attempt number in the log. pr_agent/algo/pr_processing.py [332-338] except Exception as e: get_logger().warning( - f"Failed to generate prediction with {model}", - artifact={"error": e}, + f"Failed to generate prediction with {model}{(' from deployment ' + deployment_id) if deployment_id else ''} on attempt {i + 1}/{len(all_models)}", + artifact={"error": str(e), "traceback": traceback.format_exc(), "deployment_id": deployment_id, "attempt": i + 1}, ) if i == len(all_models) - 1: # If it's the last iteration raise Exception(f"Failed to generate prediction with any model of {all_models}") from e Apply / Chat Suggestion importance[1-10]: 7 __ Why: The suggestion significantly improves error logging by making the exception serializable, adding the stack trace, and including more context, which is crucial for debugging failures.	Medium
General	Log full exception traceback Add the exception traceback to the log artifact to provide more context for debugging, which is especially helpful in a retry mechanism. pr_agent/algo/pr_processing.py [332-338] except Exception as e: get_logger().warning( f"Failed to generate prediction with {model}", - artifact={"error": e}, + artifact={"error": str(e), "traceback": traceback.format_exc()}, ) if i == len(all_models) - 1: # If it's the last iteration raise Exception(f"Failed to generate prediction with any model of {all_models}") from e Apply / Chat Suggestion importance[1-10]: 7 __ Why: This is a valuable suggestion that significantly improves debuggability by logging the full traceback, which is crucial for diagnosing failures within a retry loop.	Medium
Possible issue	Make exception log serializable Convert the raw exception object to a string before logging it in the artifact to ensure it is serializable and prevent potential logging errors. pr_agent/algo/pr_processing.py [332-338] except Exception as e: get_logger().warning( f"Failed to generate prediction with {model}", - artifact={"error": e}, + artifact={"error": str(e)}, ) if i == len(all_models) - 1: # If it's the last iteration raise Exception(f"Failed to generate prediction with any model of {all_models}") from e Apply / Chat Suggestion importance[1-10]: 6 __ Why: The suggestion correctly identifies that logging a raw exception object can cause serialization issues and proposes converting it to a string, which is a robust practice for logging.	Low
More

• Author self-review: I have reviewed the PR code suggestions, and addressed the relevant ones.

---

Previous suggestions

✅ Suggestions up to commit be93651

Category	Suggestion	Impact
Possible issue	✅ Preserve original exception context on failure Suggestion Impact: The commit updated the final raise statement to include "from e", thereby preserving the original exception context as suggested. code diff: if i == len(all_models) - 1: # If it's the last iteration - raise Exception(f"Failed to generate prediction with any model of {all_models}") + raise Exception(f"Failed to generate prediction with any model of {all_models}") from e Chain the last caught exception e when raising the final Exception to preserve the original error context for better debugging. pr_agent/algo/pr_processing.py [336-337] if i == len(all_models) - 1: # If it's the last iteration - raise Exception(f"Failed to generate prediction with any model of {all_models}") + raise Exception(f"Failed to generate prediction with any model of {all_models}") from e [Suggestion processed] Suggestion importance[1-10]: 8 __ Why: The suggestion correctly identifies that chaining the original exception e will preserve the stack trace, which is crucial for debugging why all model predictions failed.	Medium

[ofir-frd]

Thanks @yamoyamoto , approved!