apache · Yilialinn · Sep 30, 2025 · Sep 30, 2025 · Sep 30, 2025 · Sep 30, 2025
diff --git a/blog/en/blog/2025/09/30/2025-sep-monthly-report.md b/blog/en/blog/2025/09/30/2025-sep-monthly-report.md
@@ -0,0 +1,75 @@
+---
+title: "2025 Monthly Report (September 01 - September 30)"
+keywords: ["Apache APISIX", "API Gateway", "Monthly Report", "Contributor"]
+description: Our monthly Apache APISIX community report generates insights into the project's monthly developments. The reports provide a pathway into the Apache APISIX community, ensuring that you stay well-informed and actively involved.
+tags: [Community]
+image: https://static.api7.ai/uploads/2025/09/29/APa3g8ZM_sep-monthly-report-cover-en.webp
+---
+
+> Recently, we've introduced and updated some new features, including adding fallback mechanism for specific error codes in `ai-proxy-multi` plugin and adding KSUID algorithm in `request-id` plugin, etc. For more details, please read this month's newsletter.
+
+<!--truncate-->
+
+## Introduction
+
+From its inception, the Apache APISIX project has embraced the ethos of open-source community collaboration, propelling it into the ranks of the most active global open-source API gateway projects. The proverbial wisdom of 'teamwork makes the dream work' rings true in our way and is made possible by the collective effort of our community.
+
+From September 1st to September 30th, 17 contributors made 76 commits to Apache APISIX. We sincerely appreciate your contributions to Apache APISIX.
+
+## Contributor Statistics
+
+![Apache APISIX Contributors List](https://static.api7.ai/uploads/2025/09/29/8AmFT8hp_2025-sep-contributor-list.webp)
+
+![New Contributors List](https://static.api7.ai/uploads/2025/09/29/ukKJS7E3_2025-sep-new-contributors.webp)
+
+## Feature Highlights
+
+### 1. Admin API No Longer Overwrites User Data with Defaults
+
+PR: https://github.com/apache/apisix/pull/12603
+
+Contributor: [SkyeYoung](https://github.com/SkyeYoung)
+
+This PR removes the logic of populating default values before writing to the admin API. By leaving the copy in etcd untouched and applying the new defaults only inside APISIX, the PR keeps the user's config identical to what was posted, removing surprise diffs and letting the ingress controller's ADC diff logic keep working with older versions.
+
+### 2. Add the KSUID algorithm in the `request-id` Plugin
+
+PR: https://github.com/apache/apisix/pull/12573
+
+Contributor: [Crazy-xyr](https://github.com/Crazy-xyr)
+
+This PR replaces the previous request ID generation method with the KSUID (K-Sortable Unique IDentifier) algorithm. KSUIDs are always 27 characters long and are encoded in base62, making them URL-safe. Their key advantage is that they are lexicographically sortable, ensuring they remain in time order even when treated as strings. With 128 bits of randomness, they also provide stronger collision resistance than standard UUIDs.
+
+### 3. Add Fallback Mechanism for Specific Error Codes in `ai-proxy-multi` Plugin
+
+PR: https://github.com/apache/apisix/pull/12571
+
+Contributor: [Revolyssup](https://github.com/Revolyssup)
+
+The `ai-proxy-multi` plugin now retries any request that returns `429` or `5xx`, automatically failing over to the next healthy model endpoint until one succeeds, so more calls finish without surfacing an error to the client.
+
+### 4. Add Latency and Usage in Access log of `ai-proxy` Plugin and Prometheus Metrics
+
+PR: https://github.com/apache/apisix/pull/12518
+
+Contributor: [Revolyssup](https://github.com/Revolyssup)
+
+This PR adds latency and token info for the `ai-proxy` plugin in the access log for easy debugging. It also adds Prometheus metrics for AI-related requests and adds two more labels: `request_type` to distinguish between regular requests and AI-related requests; and `llm_model` for the LLM model name forwarded to the upstream LLM service.
+
+### 5. Add New ctx Variable `request_llm_model` for Requesting LLM Model
+
+PR: https://github.com/apache/apisix/pull/12554
+
+Contributor: [bzp2010](https://github.com/bzp2010)
+
+Users need to record the LLM model of client requests that are not modified by the gateway when logging AI proxy requests, so a new context variable `request_llm_model` is added to record it.
+
+## Recommended Blogs
+
+- [360 Built Unified L7 Load Balancer Using Apache APISIX](https://apisix.apache.org/blog/2025/09/03/360-built-unified-l7-load-balancer-with-apisix/)
+
+  360 unifies Layer 7 load balancing using APISIX, gaining VPC, cloud-native, and fine-grained routing in one seamless upgrade.
+
+- [Load Balancing AI/ML API with Apache APISIX](https://apisix.apache.org/blog/2025/07/31/load-balancing-between-ai-ml-api-with-apisix/)
+
+  This blog provides a step-by-step guide to configure Apache APISIX for AI traffic splitting and load balancing between API versions, covering security setup, canary testing, and deployment monitoring.
diff --git a/blog/zh/blog/2025/09/30/2025-sep-monthly-report.md b/blog/zh/blog/2025/09/30/2025-sep-monthly-report.md
@@ -0,0 +1,69 @@
+---
+title: "2025 社区月报 (09.01 - 09.30)"
+keywords: ["Apache APISIX", "API 网关", "社区月报", "贡献者"]
+description: Apache APISIX 社区的月报旨在帮助社区成员更全面地了解社区的最新动态，方便大家参与到 Apache APISIX 社区中来。
+tags: [Community]
+image: https://static.api7.ai/uploads/2025/09/29/qUyHsgwA_sep-monthly-report-cover-cn.webp
+---
+
+> 最近，我们新增并优化了一些功能，如 `request-id` 插件支持 KSUID 算法、`ai-proxy-multi` 插件针对 `429`/`5xx` 状态码增加自动重试与节点降级等。有关更多细节，请阅读本期月报。
+
+<!--truncate-->
+
+## 导语
+
+Apache APISIX 项目始终秉承着开源社区协作的精神，自问世起便崭露头角，如今已经成为全球最活跃的开源 API 网关项目之一。正如谚语所言，“众人拾柴火焰高”，这一辉煌成就，得益于整个社区伙伴的协同努力。
+
+从 2025.09.01 至 2025.09.30，有 17 名开发者提交了 76 个 commits，为 Apache APISIX 做出了重要贡献。感谢这些伙伴们对 Apache APISIX 的无私支持！正是因为你们的付出，才能让 Apache APISIX 项目不断改进、提升和壮大。
+
+## 贡献者统计
+
+![贡献者名单](https://static.api7.ai/uploads/2025/09/29/8AmFT8hp_2025-sep-contributor-list.webp)
+
+![新晋贡献者](https://static.api7.ai/uploads/2025/09/29/ukKJS7E3_2025-sep-new-contributors.webp)
+
+## 近期亮点功能
+
+### 1. Admin API 在写入时不再自动填充默认值
+
+相关 PR：https://github.com/apache/apisix/pull/12603
+
+贡献者：[SkyeYoung](https://github.com/SkyeYoung)
+
+Admin API 不再用默认值覆盖用户数据。写入 etcd 时保持原样，默认值只在 APISIX 内部补齐，既避免用户配置被悄悄改动，也保证 Ingress Controller 的 ADC 比对逻辑与旧版本兼容。
+
+### 2. `request-id` 插件支持 KSUID 算法
+
+相关 PR：https://github.com/apache/apisix/pull/12573
+
+贡献者：[Crazy-xyr](https://github.com/Crazy-xyr)
+
+本 PR 将原有的 `request-id` 生成方式替换为 KSUID（可排序唯一标识符）算法。KSUID 始终为 27 位字符，采用 base62 编码，确保其可用于 URL 场景。其核心优势在于支持字典序排序，即使作为字符串处理也能保持时间顺序。该算法还包含 128 位随机数，相比标准 UUID 具有更强的防冲突能力。
+
+### 3. `ai-proxy-multi` 插件针对 `429`/`5xx` 状态码增加自动重试与节点降级
+
+相关 PR：https://github.com/apache/apisix/pull/12571
+
+贡献者：[Revolyssup](https://github.com/Revolyssup)
+
+`ai-proxy-multi` 插件现对 `429`/`5xx` 错误自动重试，并依次切换到下一个健康模型节点，直到请求成功，减少客户端感知到的失败。
+
+### 4. `ai-proxy` 插件的访问日志与 Prometheus 指标中新增延迟和用量
+
+相关 PR：https://github.com/apache/apisix/pull/12518
+
+贡献者：[Revolyssup](https://github.com/Revolyssup)
+
+本 PR 在访问日志中为 `ai-proxy` 插件补充了延迟和 token 信息，方便排障；同时新增 AI 相关请求的 Prometheus 指标，并追加两个标签：`request_type` 用于区分普通请求与 AI 请求；`llm_model` 用于记录转发给上游 LLM 服务的模型名称。
+
+### 5. 新增上下文变量 `request_llm_model` 在日志中保留客户端原始 LLM 模型
+
+相关 PR：https://github.com/apache/apisix/pull/12554
+
+贡献者：[bzp2010](https://github.com/bzp2010)
+
+用户在记录 AI 代理请求时，需要保存未经网关修改的客户端原始 LLM 模型，因此新增上下文变量 `request_llm_model` 来存储该值。
+
+## 结语
+
+Apache APISIX 的项目[官网](https://apisix.apache.org/zh/)和 Github 上的 [Issues](https://github.com/apache/apisix/issues) 上已经积累了比较丰富的文档教程和使用经验，如果您遇到问题可以翻阅文档，用关键词在 Issues 中搜索，也可以参与 Issues 上的讨论，提出自己的想法和实践经验。