This is an automated email from the ASF dual-hosted git repository.
Aias00 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/hertzbeat.git
The following commit(s) were added to refs/heads/master by this push:
new dc10a4e564 docs(help): add push style monitor help documentation
(#1795) (#4321)
dc10a4e564 is described below
commit dc10a4e564d73f86b857e4b5ecb98beaa66be475
Author: Tanay Paul <[email protected]>
AuthorDate: Mon Aug 17 13:41:23 2026 +0530
docs(help): add push style monitor help documentation (#1795) (#4321)
Co-authored-by: aias00 <[email protected]>
---
home/docs/help/push.md | 130 +++++++++++++++++++++
.../current/help/push.md | 130 +++++++++++++++++++++
home/sidebars.json | 3 +-
3 files changed, 262 insertions(+), 1 deletion(-)
diff --git a/home/docs/help/push.md b/home/docs/help/push.md
new file mode 100644
index 0000000000..3355d7b6e6
--- /dev/null
+++ b/home/docs/help/push.md
@@ -0,0 +1,130 @@
+---
+id: push
+title: Monitoring：Push Style Monitor
+sidebar_label: Push Style Monitor
+keywords: [open source monitoring tool, open source push monitoring tool,
monitoring push metrics]
+---
+
+> HertzBeat actively collects metrics from targets on a schedule. Push Style
Monitor reverses that model — your application pushes Prometheus-format metrics
to HertzBeat, which is useful for short-lived jobs, batch processes, or
services behind firewalls.
+
+## How It Works
+
+1. HertzBeat exposes a Prometheus-compatible push endpoint.
+2. Your application POSTs metrics in **Prometheus text exposition format** to
that endpoint using a `job` name and an `instance` name.
+3. On the first push for a new `job`/`instance` pair, HertzBeat automatically
creates a monitor for it.
+4. Subsequent pushes update the metrics stored under that monitor.
+
+## Push Endpoint
+
+```http
+POST
http://{hertzbeat-host}:{port}/api/push/prometheus/job/{job}/instance/{instance}
+Content-Type: text/plain
+```
+
+| Path segment | Description | Example |
+|---|---|---|
+| `{hertzbeat-host}` | Address of the HertzBeat server | `127.0.0.1` |
+| `{port}` | HertzBeat HTTP port (default `1157`) | `1157` |
+| `{job}` | Logical name for the application (alphanumeric and `_` only) |
`my_app` |
+| `{instance}` | Instance identifier within that job (alphanumeric and `_`
only) | `server_1` |
+
+## Metrics Format
+
+The request body must follow the **Prometheus text exposition format**. Each
non-comment, non-empty line defines one sample:
+
+```promtail
+# HELP http_requests_total Total HTTP requests handled
+# TYPE http_requests_total counter
+http_requests_total{method="GET",status="200"} 1234
+http_requests_total{method="POST",status="200"} 56
+
+# HELP cpu_usage_percent Current CPU utilization
+# TYPE cpu_usage_percent gauge
+cpu_usage_percent 72.5
+
+# HELP memory_used_bytes Memory currently in use
+# TYPE memory_used_bytes gauge
+memory_used_bytes 536870912
+```
+
+## Configuration Parameters
+
+| Parameter | Description |
+|---|---|
+| Push Module Host | Address of the HertzBeat server your application will
push to. Default: `127.0.0.1` |
+| Port | HertzBeat HTTP port. Default: `1157` |
+| Metrics Fields | Define the metric field names and their types (Number /
String) that HertzBeat should expect |
+
+## Example: Shell (curl)
+
+```bash
+curl -X POST \
+ http://localhost:1157/api/push/prometheus/job/my_app/instance/server_1 \
+ -H 'Content-Type: text/plain' \
+ --data-binary @- << 'EOF'
+# HELP cpu_usage_percent Current CPU utilization
+# TYPE cpu_usage_percent gauge
+cpu_usage_percent{core="0"} 45.2
+cpu_usage_percent{core="1"} 38.7
+
+# HELP memory_used_bytes Memory currently in use
+# TYPE memory_used_bytes gauge
+memory_used_bytes 1073741824
+EOF
+```
+
+## Example: Python
+
+```python
+import requests
+
+def push_metrics(host: str, port: int, job: str, instance: str, body: str) ->
None:
+ url =
f"http://{host}:{port}/api/push/prometheus/job/{job}/instance/{instance}";
+ response = requests.post(url, data=body, headers={"Content-Type":
"text/plain"})
+ response.raise_for_status()
+
+metrics_body = """\
+# HELP request_duration_seconds Request latency
+# TYPE request_duration_seconds gauge
+request_duration_seconds{endpoint="/api/v1/users"} 0.023
+"""
+
+push_metrics("localhost", 1157, "my_app", "server_1", metrics_body)
+```
+
+## Example: Java
+
+```java
+import java.net.URI;
+import java.net.http.HttpClient;
+import java.net.http.HttpRequest;
+import java.net.http.HttpResponse;
+
+String body = """
+ # HELP jvm_memory_used_bytes JVM heap memory currently in use
+ # TYPE jvm_memory_used_bytes gauge
+ jvm_memory_used_bytes 134217728
+ """;
+
+HttpRequest request = HttpRequest.newBuilder()
+
.uri(URI.create("http://localhost:1157/api/push/prometheus/job/my_app/instance/server_1";))
+ .header("Content-Type", "text/plain")
+ .POST(HttpRequest.BodyPublishers.ofString(body))
+ .build();
+
+HttpClient.newHttpClient().send(request, HttpResponse.BodyHandlers.ofString());
+```
+
+## Common Problems
+
+1. **`Request not matched` response**
+ The `{job}` and `{instance}` path segments only accept alphanumeric
characters and underscores (`[a-zA-Z0-9_]`). Hyphens, dots, and slashes are not
accepted.
+
+2. **Metrics not appearing in the dashboard**
+ Metric field names in the HertzBeat monitor configuration must match the
metric names (or label names) in the body you are pushing exactly, including
case.
+
+3. **Push rejected with no monitor created**
+ HertzBeat caps the number of push monitors it will auto-create (default:
10,000). If the cap is reached, pushes from unknown `job`/`instance` pairs are
rejected. Existing monitors continue to receive data normally.
+
+4. **Body too large**
+ Single push requests are limited to 5 MB and 10,000 samples by default.
Split large payloads across multiple requests.
diff --git
a/home/i18n/zh-cn/docusaurus-plugin-content-docs/current/help/push.md
b/home/i18n/zh-cn/docusaurus-plugin-content-docs/current/help/push.md
new file mode 100644
index 0000000000..8ffdd385f7
--- /dev/null
+++ b/home/i18n/zh-cn/docusaurus-plugin-content-docs/current/help/push.md
@@ -0,0 +1,130 @@
+---
+id: push
+title: 监控：推送方式监控
+sidebar_label: 推送方式监控
+keywords: [开源监控系统, 开源推送监控, 推送方式指标监控]
+---
+
+> HertzBeat 默认以主动采集方式定时拉取目标指标。推送方式监控反转了这一模型——由您的应用将 Prometheus 格式的指标推送至
HertzBeat，适用于短生命周期任务、批处理作业或位于防火墙内部的服务。
+
+## 工作原理
+
+1. HertzBeat 暴露一个兼容 Prometheus Pushgateway 协议的推送端点。
+2. 您的应用以 **Prometheus 文本格式** 将指标 POST 到该端点，需指定 `job` 名称和 `instance` 名称。
+3. 对于首次推送的新 `job`/`instance` 组合，HertzBeat 将自动为其创建一个监控实例。
+4. 后续推送将持续更新该监控实例下存储的指标数据。
+
+## 推送端点
+
+```http
+POST
http://{hertzbeat-host}:{port}/api/push/prometheus/job/{job}/instance/{instance}
+Content-Type: text/plain
+```
+
+| 路径参数 | 说明 | 示例 |
+|---|---|---|
+| `{hertzbeat-host}` | HertzBeat 服务器地址 | `127.0.0.1` |
+| `{port}` | HertzBeat HTTP 端口（默认 `1157`） | `1157` |
+| `{job}` | 应用逻辑名称（仅允许字母、数字和下划线 `_`） | `my_app` |
+| `{instance}` | 该 job 下的实例标识（仅允许字母、数字和下划线 `_`） | `server_1` |
+
+## 指标格式
+
+请求体须遵循 **Prometheus 文本格式**，每行（非注释、非空行）定义一个采样点：
+
+```promtail
+# HELP http_requests_total 处理的 HTTP 请求总数
+# TYPE http_requests_total counter
+http_requests_total{method="GET",status="200"} 1234
+http_requests_total{method="POST",status="200"} 56
+
+# HELP cpu_usage_percent 当前 CPU 使用率
+# TYPE cpu_usage_percent gauge
+cpu_usage_percent 72.5
+
+# HELP memory_used_bytes 当前内存使用量
+# TYPE memory_used_bytes gauge
+memory_used_bytes 536870912
+```
+
+## 配置参数
+
+| 参数名称 | 参数帮助描述 |
+|---|---|
+| 推送模块 Host | 您的应用将指标推送到的 HertzBeat 服务器地址，默认：`127.0.0.1` |
+| 端口 | HertzBeat HTTP 端口，默认：`1157` |
+| 监控数据字段 | 定义 HertzBeat 需要接收的指标字段名及其类型（数值 / 字符串） |
+
+## 示例：Shell（curl）
+
+```bash
+curl -X POST \
+ http://localhost:1157/api/push/prometheus/job/my_app/instance/server_1 \
+ -H 'Content-Type: text/plain' \
+ --data-binary @- << 'EOF'
+# HELP cpu_usage_percent 当前 CPU 使用率
+# TYPE cpu_usage_percent gauge
+cpu_usage_percent{core="0"} 45.2
+cpu_usage_percent{core="1"} 38.7
+
+# HELP memory_used_bytes 当前内存使用量
+# TYPE memory_used_bytes gauge
+memory_used_bytes 1073741824
+EOF
+```
+
+## 示例：Python
+
+```python
+import requests
+
+def push_metrics(host: str, port: int, job: str, instance: str, body: str) ->
None:
+ url =
f"http://{host}:{port}/api/push/prometheus/job/{job}/instance/{instance}";
+ response = requests.post(url, data=body, headers={"Content-Type":
"text/plain"})
+ response.raise_for_status()
+
+metrics_body = """\
+# HELP request_duration_seconds 请求耗时
+# TYPE request_duration_seconds gauge
+request_duration_seconds{endpoint="/api/v1/users"} 0.023
+"""
+
+push_metrics("localhost", 1157, "my_app", "server_1", metrics_body)
+```
+
+## 示例：Java
+
+```java
+import java.net.URI;
+import java.net.http.HttpClient;
+import java.net.http.HttpRequest;
+import java.net.http.HttpResponse;
+
+String body = """
+ # HELP jvm_memory_used_bytes JVM 堆内存当前使用量
+ # TYPE jvm_memory_used_bytes gauge
+ jvm_memory_used_bytes 134217728
+ """;
+
+HttpRequest request = HttpRequest.newBuilder()
+
.uri(URI.create("http://localhost:1157/api/push/prometheus/job/my_app/instance/server_1";))
+ .header("Content-Type", "text/plain")
+ .POST(HttpRequest.BodyPublishers.ofString(body))
+ .build();
+
+HttpClient.newHttpClient().send(request, HttpResponse.BodyHandlers.ofString());
+```
+
+## 常见问题
+
+1. **响应提示 `Request not matched`**
+ `{job}` 和 `{instance}` 路径参数仅允许字母、数字和下划线（`[a-zA-Z0-9_]`），不支持连字符、点号或斜杠。
+
+2. **指标推送成功但仪表盘未显示数据**
+ HertzBeat 监控实例中配置的字段名称须与推送体中的指标名称或标签名称完全一致（区分大小写）。
+
+3. **推送被拒绝且未自动创建监控**
+ HertzBeat 对可自动创建的推送监控数量有上限（默认 10,000 个）。达到上限后，来自未知 `job`/`instance`
组合的推送将被拒绝，但已存在的监控可正常接收数据。
+
+4. **请求体过大**
+ 单次推送请求默认限制为 5 MB 且最多 10,000 个采样点，请将较大的数据载荷拆分为多次请求发送。
diff --git a/home/sidebars.json b/home/sidebars.json
index b42a8db679..1bfe5c2e3d 100755
--- a/home/sidebars.json
+++ b/home/sidebars.json
@@ -159,7 +159,8 @@
"help/websocket",
"help/mqtt",
"help/modbus",
- "help/jenkins"
+ "help/jenkins",
+ "help/push"
]
},
{
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
