Feat/reuse http connection 2 #105
Conversation
This reverts commit 9f529ec.
Summary of ChangesHello @kevinlin09, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request refactors the HTTP request handling in the DashScope SDK to enable HTTP connection reuse and custom session support. It removes the global session management in favor of users providing their own sessions, offering greater flexibility and control over HTTP connection configurations. The changes include modifications to the Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here. You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension. Footnotes
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request refactors HTTP connection management by removing global session managers, allowing users to provide their own requests.Session or aiohttp.ClientSession objects for improved resource management and dependency injection. However, it introduces a significant resource leak in the asynchronous request handling due to unclosed aiohttp connectors, which could lead to Denial of Service under high load. While the changes offer a cleaner and more flexible API, are well-supported by tests, and include updated examples, there is also a suggestion for improving code robustness in http_request.py.
| connector = aiohttp.TCPConnector( | ||
| ssl=ssl.create_default_context( | ||
| cafile=certifi.where(), | ||
| ), | ||
| ) | ||
| session = aiohttp.ClientSession( | ||
| connector=connector, | ||
| timeout=aiohttp.ClientTimeout(total=self.timeout), | ||
| headers=self.headers, | ||
| ) | ||
| should_close = True |
There was a problem hiding this comment.
A new aiohttp.TCPConnector is created here for every request that doesn't use an external session. However, this connector is not explicitly closed, and aiohttp.ClientSession does not take ownership of connectors passed to it by default. This leads to a resource leak of file descriptors and sockets for every asynchronous request made. Over time, this can lead to resource exhaustion and a Denial of Service (DoS) condition in long-running applications. Additionally, the current error handling in the finally block may raise an UnboundLocalError if an exception occurs during session initialization, as session and should_close might not be defined.
| connector = aiohttp.TCPConnector( | |
| ssl=ssl.create_default_context( | |
| cafile=certifi.where(), | |
| ), | |
| ) | |
| session = aiohttp.ClientSession( | |
| connector=connector, | |
| timeout=aiohttp.ClientTimeout(total=self.timeout), | |
| headers=self.headers, | |
| ) | |
| should_close = True | |
| connector = aiohttp.TCPConnector( | |
| ssl=ssl.create_default_context( | |
| cafile=certifi.where(), | |
| ), | |
| ) | |
| session = aiohttp.ClientSession( | |
| connector=connector, | |
| connector_owner=True, | |
| timeout=aiohttp.ClientTimeout(total=self.timeout), | |
| headers=self.headers, | |
| ) | |
| should_close = True |
| if ( | ||
| session_type == "ClientSession" and "aiohttp" in session_module | ||
| ) or isinstance(session, aiohttp.ClientSession): | ||
| self._external_session = None | ||
| self._external_aio_session = session | ||
| else: | ||
| # Treat as requests Session | ||
| self._external_session = session | ||
| self._external_aio_session = None |
There was a problem hiding this comment.
The logic for detecting the session type is good, but it could be more robust. The if block explicitly checks for an aiohttp.ClientSession, but the else block implicitly assumes any other object is a requests.Session. This could lead to runtime errors if an unsupported session type is passed. It would be safer to add an explicit check for requests.Session and raise a TypeError for any other type.
if (
session_type == "ClientSession" and "aiohttp" in session_module
) or isinstance(session, aiohttp.ClientSession):
self._external_session = None
self._external_aio_session = session
elif isinstance(session, requests.Session):
# Treat as requests Session
self._external_session = session
self._external_aio_session = None
else:
raise TypeError(
'Unsupported session type: %s. '
'Expected requests.Session or aiohttp.ClientSession.' %
type(session))1 participant
Description
[Describe what this PR does and why]
Related Issue: Fixes #[issue_number] or Relates to #[issue_number]
Security Considerations: [Check if API keys or sensitive credentials are exposed in code/logs]
Type of Change
Component(s) Affected
Checklist
Testing
[How to test these changes]
Additional Notes
[Optional: any other context]