docs(mongodb): re-verify Percona MongoDB Operator guide on ACP 4.4

docs/en/solutions/How_to_Deploy_Community_Percona_MongoDB_Operator.md

@@ -14,7 +14,7 @@ id: KB260400014
 
 This guide walks you through deploying MongoDB on Alauda Container Platform using the upstream **community** [Percona Server for MongoDB Operator](https://github.com/percona/percona-server-mongodb-operator). The previously-bundled Alauda MongoDB plugin is no longer distributed via the ACP marketplace, so this guide provides a self-service path using the community release.
 
-**Verified versions** (verified on ACP 4.2 / Kubernetes 1.33; check the upstream documentation for newer releases):
+**Verified versions** (verified on ACP 4.2 / Kubernetes 1.33 and ACP 4.4 / Kubernetes 1.35; check the upstream documentation for newer releases):
 
 | Component | Version |
 | :--- | :--- |
@@ -106,6 +106,14 @@ done
 
 ### Using the ACP-integrated Harbor registry
 
+If you are not sure which registry address your cluster uses, read it from the platform:
+
+```bash
+kubectl -n kube-public get configmap global-info -o jsonpath='{.data.registryAddress}'
+```
+
+Use that value (host and port) as `REGISTRY_SERVER` below instead of the `:45443` example.
+
 If you are pushing to the Harbor registry that ships with ACP (typical endpoint `https://<acp-portal-host>:45443`), the push credentials live in a Secret on the management cluster — they are not the same as your ACP portal login:
 
 ```bash
@@ -339,7 +347,7 @@ For the full Custom Resource field reference and all available options (TLS, mon
 kubectl -n "$NS" get psmdb -w
 ```
 
-Wait for `STATUS=ready`. On healthy storage, the cluster reaches ready within ~60 seconds.
+Wait for `STATUS=ready`. On healthy storage, the cluster reaches ready within about 1–2 minutes. It is normal for `STATUS` to show `error` or `initializing` briefly while the operator generates the cluster's TLS secrets and starts the pods; only a status that stays `error` needs investigation (check the operator log).
 
 ```text
 NAME       ENDPOINT                                            STATUS   AGE
@@ -384,7 +392,7 @@ This guide is scoped to a baseline deployment: install the operator, wire it to
 
 To set expectations clearly:
 
-- **Verified by Alauda** features have been tested end-to-end on a representative ACP cluster (ACP 4.2 / Kubernetes 1.33, operator v1.22.0). They work as documented here.
+- **Verified by Alauda** features have been tested end-to-end on a representative ACP cluster (ACP 4.2 / Kubernetes 1.33, operator v1.22.0). They work as documented here. The baseline deployment in this guide (namespace-scoped operator install from a private registry, sharded MongoDB `8.0.19-7` cluster reaching `ready`, and the Step 6 access check) was re-verified on ACP 4.4 / Kubernetes 1.35 with operator v1.22.0.
 - **Not verified by Alauda** features may work, but Alauda has not tested them on ACP. If your use case depends on them, treat the upstream Percona documentation as authoritative and validate in your own environment before relying on them in production.
 
 ### Verified by Alauda

docs/zh/solutions/How_to_Deploy_Community_Percona_MongoDB_Operator.md

@@ -15,7 +15,7 @@ sourceSHA: 303099030817594e33348262986ed3861e759e0dce0c1dbf85c7f5fa12869df9
 
 本指南将引导您在 Alauda 容器平台上使用上游 **community** [Percona Server for MongoDB Operator](https://github.com/percona/percona-server-mongodb-operator) 部署 MongoDB。之前捆绑的 Alauda MongoDB 插件不再通过 ACP 市场分发，因此本指南提供了使用社区版本的自助路径。
 
-**验证版本**（在 ACP 4.2 / Kubernetes 1.33 上验证；请查看上游文档以获取更新版本）：
+**验证版本**（在 ACP 4.2 / Kubernetes 1.33 和 ACP 4.4 / Kubernetes 1.35 上验证；请查看上游文档以获取更新版本）：
 
 | 组件                                 | 版本                  |
 | :---------------------------------- | :-------------------- |
@@ -107,6 +107,14 @@ done
 
 ### 使用 ACP 集成的 Harbor 注册表
 
+如果不确定集群使用的镜像仓库地址，可以从平台读取：
+
+```bash
+kubectl -n kube-public get configmap global-info -o jsonpath='{.data.registryAddress}'
+```
+
+将该值（主机和端口）作为下面的 `REGISTRY_SERVER`，替代 `:45443` 示例。
+
 如果您要推送到随 ACP 提供的 Harbor 注册表（典型端点 `https://<acp-portal-host>:45443`），则推送凭据存储在管理集群的 Secret 中 — 它们与您的 ACP 门户登录不同：
 
 ```bash
@@ -338,7 +346,7 @@ EOF
 kubectl -n "$NS" get psmdb -w
 ```
 
-等待 `STATUS=ready`。在健康的存储上，集群在 ~60 秒内达到就绪状态。
+等待 `STATUS=ready`。在健康的存储上，集群约在 1–2 分钟内达到就绪状态。在 operator 生成集群 TLS Secret 并启动 Pod 期间，`STATUS` 短暂显示 `error` 或 `initializing` 属于正常现象；只有持续停留在 `error` 时才需要排查（查看 operator 日志）。
 
 ```text
 NAME       ENDPOINT                                            STATUS   AGE
@@ -383,7 +391,7 @@ kubectl -n "$NS" port-forward "svc/${CLUSTER}-mongos" 27017:27017
 
 为了清晰地设定期望：
 
-- **Alauda 验证的**功能已在代表性的 ACP 集群（ACP 4.2 / Kubernetes 1.33，operator v1.22.0）上进行了端到端测试。它们按照此处记录的方式工作。
+- **Alauda 验证的**功能已在代表性的 ACP 集群（ACP 4.2 / Kubernetes 1.33，operator v1.22.0）上进行了端到端测试。它们按照此处记录的方式工作。本指南的基础部署流程（从私有镜像仓库以命名空间范围安装 operator、分片 MongoDB `8.0.19-7` 集群达到 `ready`，以及步骤 6 的访问检查）已在 ACP 4.4 / Kubernetes 1.35、operator v1.22.0 上重新验证。
 - **未经过 Alauda 验证的**功能可能有效，但 Alauda 尚未在 ACP 上测试它们。如果您的用例依赖于它们，请将上游 Percona 文档视为权威，并在生产环境中依赖之前在您自己的环境中进行验证。
 
 ### Alauda 验证的