Hacktoberfest 2026：メンテナが10月に向けて印を付けた、オープンで初心者向けの issue。 Hacktoberfest の issue を見る

[aio] host_monitoring_v2 without a topology plugin fails the first statement on cluster-endpoint connections

オープン
  • 何か問題がありますか？
#1,288 コメント 0 件 リアクション 0 件 担当者 0 名 GitHub で見る

メンテナーはふだん 1 日以内に返信

@AhmadMasry がすでに取り組んでいます。

2026年10月6日 から。

  • #1290 @AhmadMasry による — オープン

評価

難易度
4/5
見積もり時間
3〜5日
初心者へのやさしさ
49/100
issue の種類
バグ
明瞭さ
おおむね明確
活発さ
活発
技術スタック
postgresql, python, sqlalchemy
領域
backend, databases

調査の方向性

Start with aio/wrapper.py and _build_host_list_provider, then compare the Aurora dialect provider selection in database_dialect.py with the async behavior. Trace how AsyncHostMonitoringPlugin._get_monitoring_host_info identifies and caches a host, using the described mocked reproduction as a starting point. Done when the first statement succeeds through a cluster endpoint and monitoring targets the connected instance, with regression coverage for the failure and cache behavior.

索引モデルが issue の本文から書いたものです。

説明

Describe the bug

With the async API, host_monitoring_v2 used without a topology plugin (no failover, failover_v2, aurora_connection_tracker, ...) fails the first statement on every new connection made through an Aurora cluster endpoint, with AwsWrapperError: [HostMonitoringV2Plugin] Unable to identify the connected database instance. Later statements on that connection succeed, but the monitor watches the cluster endpoint's DNS name instead of the instance the connection is on.

Cause:

  • aio/wrapper.py _build_host_list_provider returns AsyncStaticHostListProvider when plugins contains none of _TOPOLOGY_REQUIRING_PLUGINS, whatever the database dialect. host_monitoring_v2 isn't in that set.
  • For a cluster endpoint, AsyncHostMonitoringPlugin._get_monitoring_host_info calls plugin_service.identify_connection. With the static provider the topology has no host_id, so it returns None and the plugin raises before running the statement.
  • _get_monitoring_host_info assigns self._monitoring_host_info = current_host_info before identifying, so on the next statement it returns the cached cluster endpoint and monitors that.

The sync wrapper doesn't hit this: Aurora dialects always use RdsHostListProvider (database_dialect.py, AuroraPgDialect.get_host_list_provider_supplier), regardless of plugins, so identify_connection resolves the instance.

Expected Behavior

As in sync, host_monitoring_v2 alone on a cluster endpoint identifies the instance the connection landed on and monitors it, and the first statement runs normally.

What plugins are used? What other connection properties were set?

wrapper_dialect=aurora-pg&wrapper_plugins=host_monitoring_v2, connecting through the cluster (writer) endpoint. With failover added, or through an instance endpoint, it works.

Current Behavior

Checked without a database by building the provider the way connect does and calling _get_monitoring_host_info with get_instance_id patched to return an instance id:

AsyncStaticHostListProvider  mydb.cluster-xyz...  -> 1st call: AwsWrapperError: Unable to identify the connected database instance
                                                     2nd call: monitors mydb.cluster-xyz.us-east-1.rds.amazonaws.com
AsyncStaticHostListProvider  inst-1.xyz...        -> monitors inst-1.xyz.us-east-1.rds.amazonaws.com
Reproduction Steps
import asyncio
from sqlalchemy import text
from sqlalchemy.ext.asyncio import create_async_engine
from aws_advanced_python_wrapper.aio import release_resources_async


async def main():
    engine = create_async_engine(
        "postgresql+aws_wrapper_psycopg://USER:PASSWORD@<name>.cluster-<id>.<region>.rds.amazonaws.com:5432/DB"
        "?wrapper_dialect=aurora-pg&wrapper_plugins=host_monitoring_v2")
    try:
        async with engine.connect() as conn:
            await conn.execute(text("SELECT 1"))  # raises AwsWrapperError
    finally:
        await engine.dispose()
        await release_resources_async()

asyncio.run(main())
Possible Solution

Match sync: pick the topology-aware provider from the Aurora/Multi-AZ/Global dialect even when no topology plugin is configured, or add host_monitoring_v2 (and host_monitoring) to _TOPOLOGY_REQUIRING_PLUGINS. Either way, resolve the instance (the topology monitor is shared per cluster after #1285, so this costs one monitoring connection per cluster). Separately, _get_monitoring_host_info shouldn't cache the cluster endpoint when identification fails.

Additional Information/Context

Found while answering a user question about using host_monitoring_v2 without failover. Related: #1284.

The AWS Advanced Python Wrapper version used

main at 8feea85 (3.1.0)

python version used

Python 3.14

Operating System and version

macOS (checked locally, not against a live cluster)

主要言語
Python
スター
99
フォーク
22
平均マージ
1日 7時間
マージ済み PR（30日）
4

環境構築

はじめの一歩

  1. issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
  2. 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
  3. リポジトリをフォークし、ブランチを切って変更します。
  4. issue 番号を参照したプルリクエストを送ります。

aws/aws-advanced-python-wrapper のほかの issue

aws/aws-advanced-python-wrapper の issue をすべて見る

似ている issue

Python の issue をもっと見る

新しい issue をメールで受け取る

初心者向けの GitHub issue を短くまとめたダイジェスト。