Skip to content

Commit e50f58f

Browse files
alick-limingalick-liming
authored
Merge pull request #328 from flashcatcloud/codex/huawei-aom-integration-docs-20260901
Codex/huawei aom integration docs 20260901
2 parents e54bc94 + dea2c33 commit e50f58f

4 files changed

Lines changed: 575 additions & 0 deletions

File tree

‎docs.json‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1681,6 +1681,7 @@
16811681
"zh/on-call/integration/alert-integration/alert-sources/huawei-ces",
16821682
"zh/on-call/integration/alert-integration/alert-sources/huawei-lts",
16831683
"zh/on-call/integration/alert-integration/alert-sources/huawei-apm",
1684+
"zh/on-call/integration/alert-integration/alert-sources/huawei-aom",
16841685
"zh/on-call/integration/alert-integration/alert-sources/volcengine-cm-metrics",
16851686
"zh/on-call/integration/alert-integration/alert-sources/volcengine-cm-event",
16861687
"zh/on-call/integration/alert-integration/alert-sources/volcengine-tls",
@@ -3058,6 +3059,7 @@
30583059
"en/on-call/integration/alert-integration/alert-sources/huawei-ces",
30593060
"en/on-call/integration/alert-integration/alert-sources/huawei-lts",
30603061
"en/on-call/integration/alert-integration/alert-sources/huawei-apm",
3062+
"en/on-call/integration/alert-integration/alert-sources/huawei-aom",
30613063
"en/on-call/integration/alert-integration/alert-sources/volcengine-cm-metrics",
30623064
"en/on-call/integration/alert-integration/alert-sources/volcengine-cm-event",
30633065
"en/on-call/integration/alert-integration/alert-sources/volcengine-tls",

‎en/on-call/integration/alert-integration/alert-sources/huawei-aom.mdx‎

Lines changed: 286 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,286 @@
1+
---
2+
title: "Huawei Cloud AOM alert integration"
3+
description: "Send Huawei Cloud AOM alerts to Flashduty On-call through SMN notifications or AOM webhook data subscriptions"
4+
keywords: ["alert integration", "Huawei Cloud AOM", "SMN", "webhook", "data ingestion"]
5+
---
6+
7+
Huawei Cloud Application Operations Management (AOM) can send triggered and recovered alerts to Flashduty On-call through Simple Message Notification (SMN) or webhook data subscriptions. Both methods use the same AOM integration push URL.
8+
9+
<Info>
10+
Use **SMN notifications** for most accounts. AOM webhook data subscription is not generally available and should only be used when Huawei Cloud has enabled it for your account.
11+
</Info>
12+
13+
<div className="hide">
14+
15+
## Create an integration in Flashduty On-call
16+
17+
Use either of the following methods to obtain an integration push URL.
18+
19+
### Use a dedicated integration
20+
21+
Use this method when you do not need to route alerts to different channels.
22+
23+
<AccordionGroup>
24+
<Accordion title="Expand">
25+
1. In the Flashduty console, select **Channels** and open the target channel.
26+
2. Select **Integrations** and click **Add integration**.
27+
3. Select **Huawei Cloud AOM** and click **Save**.
28+
4. Open the generated integration card and copy the **push URL**.
29+
</Accordion>
30+
</AccordionGroup>
31+
32+
### Use a shared integration
33+
34+
Use this method when you need to route alerts to different channels based on their payload.
35+
36+
<AccordionGroup>
37+
<Accordion title="Expand">
38+
1. In the Flashduty console, select **Integration Center > Alert events**.
39+
2. Select **Huawei Cloud AOM** and enter an integration name.
40+
3. Configure the default route and select a channel. You can add more routing rules under **Routes** after creating the integration.
41+
4. Click **Save** and copy the generated **push URL**.
42+
</Accordion>
43+
</AccordionGroup>
44+
45+
</div>
46+
47+
## Prerequisites
48+
49+
- You have created a **Huawei Cloud AOM** integration in Flashduty and obtained its HTTPS push URL.
50+
- You have created or prepared an AOM alarm rule.
51+
- To use SMN, your Huawei Cloud account can create topics, subscribe endpoints, and list topics.
52+
- The alarm rule sends notifications for both **alarm triggered** and **alarm recovered**.
53+
54+
## Method 1: Use SMN
55+
56+
### Create an SMN topic and subscription
57+
58+
<Steps>
59+
<Step title="Create an SMN topic">
60+
1. Sign in to the Huawei Cloud console and open **Simple Message Notification (SMN)**.
61+
2. Select **Topic Management > Topics** and click **Create Topic**.
62+
3. Set the topic name to **Flashduty** and save it.
63+
64+
<Frame>
65+
<img src="https://download.flashcat.cloud/flashduty/doc/en/hw/lts-1.png" alt="Create a Flashduty topic in Huawei Cloud SMN" />
66+
</Frame>
67+
</Step>
68+
69+
<Step title="Add an HTTPS subscription">
70+
1. Open the Flashduty topic and click **Add Subscription**.
71+
2. Select **HTTPS** as the protocol.
72+
3. Enter the Flashduty AOM integration **push URL** as the endpoint and save the subscription.
73+
74+
<Frame>
75+
<img src="https://download.flashcat.cloud/flashduty/doc/en/hw/lts-3.png" alt="Add the Flashduty HTTPS endpoint to an SMN topic" />
76+
</Frame>
77+
78+
Flashduty verifies the SMN confirmation request and automatically visits the Huawei Cloud confirmation URL. SMN starts sending alerts after the subscription status changes to **Confirmed**.
79+
</Step>
80+
</Steps>
81+
82+
### Configure an AOM message template
83+
84+
In the AOM 2.0 console, go to **Alarm Center > Alarm Notifications > Message Templates**. In some console versions, the path is **Alarms > Alarm Action Rules > Message Templates**.
85+
86+
When creating the template, select **HTTP/HTTPS** as the notification method. Flashduty supports the following content formats:
87+
88+
<Tabs>
89+
<Tab title="HTML">
90+
HTML templates can be in Chinese or English. Add the report type as the first line so that Flashduty can reliably distinguish triggered and recovered alerts.
91+
92+
**Chinese template**
93+
94+
~~~text
95+
上报类型：${clear_type};
96+
通知类型：${event_type};
97+
事件级别：${event_severity};
98+
事件名称：${event_name_alias};
99+
发生时间：${starts_at};
100+
事件源：$event.metadata.resource_provider;
101+
资源类型：$event.metadata.resource_type;
102+
资源标识：${resources_new};
103+
可能原因：${alarm_probableCause_zh};
104+
附加信息：${message};
105+
修复建议：${alarm_fix_suggestion_zh};
106+
~~~
107+
108+
**English template**
109+
110+
~~~text
111+
Clear Type:${clear_type};
112+
Alarm Name:${event_name_alias};
113+
Alarm ID:${id};
114+
Notification Rule:${action_rule};
115+
Occurred:${starts_at};
116+
Event Severity:${event_severity};
117+
Alarm Info:${alarm_info};
118+
Resource Identifier:${resources_new};
119+
Suggestion:${alarm_fix_suggestion_zh};
120+
~~~
121+
122+
<Tip>
123+
Put each field on a separate line and keep the trailing semicolon. A resource identifier can span multiple lines. Flashduty parses its key-value pairs into labels.
124+
</Tip>
125+
</Tab>
126+
127+
<Tab title="JSON">
128+
After selecting **JSON**, keep the default JSON content generated by AOM. Flashduty parses this JSON from the <code>message</code> field in the outer SMN notification.
129+
130+
The default AOM JSON contains the following core fields. Keep every field marked as required. Flashduty uses optional fields for labels or recovery detection when they have a value.
131+
132+
| Field | Required | Purpose |
133+
| --- | --- | --- |
134+
| `id` | No | AOM event ID; stored as the `event_id` label when present |
135+
| `starts_at` | Yes | Trigger time as a Unix timestamp in milliseconds |
136+
| `ends_at` | No | A value greater than 0 indicates recovery; a missing, empty, or 0 value falls back to the SMN subject |
137+
| `metadata.event_name` | Yes | Alert name |
138+
| `metadata.event_severity` | Yes | Alert severity |
139+
| `metadata.event_type` | Yes | Event type, usually `alarm` |
140+
| `metadata.resource_provider` | Yes | Event source, such as `AOM` or `CCE` |
141+
| `metadata.resource_type` | Yes | Resource type |
142+
| `metadata.resource_id` | Yes | Resource identifier |
143+
144+
<Warning>
145+
Do not manually add SMN envelope fields such as <code>signature</code>, <code>subject</code>, or <code>topic_urn</code> to the AOM JSON template. SMN generates the outer notification and its signature.
146+
</Warning>
147+
</Tab>
148+
</Tabs>
149+
150+
### Create an alarm notification rule
151+
152+
<Steps>
153+
<Step title="Associate the SMN topic and message template">
154+
1. In the AOM 2.0 console, go to **Alarm Center > Alarm Notifications**.
155+
2. Under **Notification Rule Management**, create an alarm notification rule.
156+
3. Select the **Flashduty** SMN topic and the AOM message template created earlier.
157+
</Step>
158+
159+
<Step title="Send triggered and recovered notifications">
160+
Enable notifications in the AOM alarm rule and make sure that the notification scenarios include both **alarm triggered** and **alarm recovered**.
161+
</Step>
162+
163+
<Step title="Verify the integration">
164+
Trigger a test alert and then recover it. In Flashduty, verify that the integration's latest event time has changed and that the alert details contain the expected severity, resource, and recovery status.
165+
</Step>
166+
</Steps>
167+
168+
## Method 2: Use an AOM webhook data subscription
169+
170+
<Warning>
171+
AOM webhook data subscription is not generally available. If the webhook alarm subscription option is not available in your console, submit a Huawei Cloud service ticket or use SMN instead.
172+
</Warning>
173+
174+
If Huawei Cloud has enabled this feature for your account, go to **Settings > Global Settings > Data Subscription** in the AOM 2.0 console. Create a webhook alarm subscription and set its destination to the same Flashduty AOM integration push URL.
175+
176+
The webhook request must use the official AOM structured event format:
177+
178+
~~~json
179+
{
180+
"events": [
181+
{
182+
"id": "4346299651651991683",
183+
"starts_at": 1664001830491,
184+
"ends_at": 0,
185+
"timeout": 300000,
186+
"resource_group_id": "project-id",
187+
"metadata": {
188+
"event_name": "Pod memory OOM",
189+
"event_severity": "Major",
190+
"event_type": "alarm",
191+
"resource_provider": "CCE",
192+
"resource_type": "service",
193+
"resource_id": "clusterName=production;kind=Pod;namespace=default;name=example;uid=example-uid"
194+
},
195+
"annotations": {
196+
"alarm_probableCause_en_us": "The pod does not have enough memory",
197+
"message": "pod was OOM killed",
198+
"alarm_fix_suggestion_en_us": "Increase the container memory limit"
199+
}
200+
}
201+
],
202+
"project_id": "project-id"
203+
}
204+
~~~
205+
206+
A request can contain up to 100 events. Each structured event must contain a valid <code>id</code>, <code>starts_at</code>, and all six required <code>metadata</code> fields.
207+
208+
## Correlate triggered and recovered alerts
209+
210+
Flashduty detects recovery in the following order:
211+
212+
| Input format | Recovery detection |
213+
| --- | --- |
214+
| SMN HTML | Read `上报类型` or `Clear Type` first; if the value is missing or unknown, inspect the SMN `subject` |
215+
| SMN JSON | `ends_at > 0` indicates recovery; if it is missing, empty, or 0, inspect the SMN `subject` |
216+
| Webhook data subscription | `ends_at > 0` indicates recovery |
217+
218+
For Huawei Cloud message templates, the Chinese `clear_type` values are `触发` and `恢复`, and the English values are `generated` and `cleared`. Flashduty uses these values first to distinguish triggered and recovered alerts.
219+
220+
<Note>
221+
If an SMN HTML message does not contain a recognized report type, a subject containing <code>恢复</code>, <code>_resolved</code>, <code>_recovered</code>, or <code>_clear</code> is treated as a recovery.
222+
</Note>
223+
224+
The following fields must remain the same between a triggered alert and its recovery:
225+
226+
- <code>metadata.event_type</code>
227+
- <code>metadata.event_name</code>
228+
- <code>metadata.resource_provider</code>
229+
- <code>metadata.resource_type</code>
230+
- <code>metadata.resource_id</code>
231+
232+
Flashduty generates the alert fingerprint from these fields. The <code>id</code> is stored as a label but does not participate in correlation, so triggered and recovered notifications may have different IDs.
233+
234+
If two independent alerts have identical values for all five fields, Flashduty treats them as the same alert. Make sure that <code>metadata.event_name</code> or <code>metadata.resource_id</code> distinguishes alerts that must be handled separately.
235+
236+
## Field and severity mappings
237+
238+
Common AOM fields are converted into Flashduty labels:
239+
240+
| AOM field | Flashduty label |
241+
| --- | --- |
242+
| `id` | `event_id` |
243+
| `metadata.event_name` | `event_name` and `check` |
244+
| `metadata.event_severity` | `event_severity` |
245+
| `metadata.resource_provider` | `resource_provider` |
246+
| `metadata.resource_type` | `resource_type` |
247+
| `metadata.resource_id` | `resource_id`, plus labels parsed from its resource key-value pairs |
248+
| `policy.alarm_rule_name` | `alarm_rule_name` |
249+
| `annotations.tms_tags` | `tms_tags` |
250+
| `annotations.generator_url` | `generator_url` |
251+
| `Notification Rule` in the English HTML template | `notification_rule` |
252+
253+
Severity mapping:
254+
255+
| AOM severity | Flashduty severity | Status |
256+
| --- | --- | --- |
257+
| Critical, Major, 紧急, or 重要 | Critical | Critical |
258+
| Minor, Warning, 次要, or 警告 | Warning | Warning |
259+
| Info or 提示 | Info | Info |
260+
261+
## Troubleshooting
262+
263+
<AccordionGroup>
264+
<Accordion title="Why does the SMN subscription remain unconfirmed?">
265+
Make sure that the endpoint is the complete HTTPS integration push URL and that its <code>integration_key</code> has not been changed. Flashduty only visits valid Huawei Cloud HTTPS confirmation URLs.
266+
</Accordion>
267+
268+
<Accordion title="Why does a manually replayed SMN request fail signature verification?">
269+
The SMN <code>signature</code> is bound to the message body, message ID, <code>subject</code>, topic, timestamp, and message type. Changing any of these fields invalidates the original signature. Send a real test notification through AOM and SMN instead of modifying a historical request.
270+
</Accordion>
271+
272+
<Accordion title="Why did the recovery not close the alert?">
273+
Make sure that the alarm rule sends recovery notifications. Add <code>$&#123;clear_type&#125;</code> to an HTML template. For SMN JSON, Flashduty first checks for an <code>ends_at</code> value greater than 0 and falls back to the <code>subject</code> when the value is missing, empty, or 0. A webhook data subscription must provide an <code>ends_at</code> value greater than 0. Also verify that the five correlation fields are identical in the triggered and recovered events.
274+
</Accordion>
275+
276+
<Accordion title="What is the difference between SMN JSON and a webhook data subscription?">
277+
SMN JSON is carried in the <code>message</code> field of an SMN notification, whose outer object still contains a signature and subject. A webhook data subscription directly sends an <code>&#123;"events":[...]&#125;</code> object without SMN fields. Do not combine both outer formats in one request.
278+
</Accordion>
279+
</AccordionGroup>
280+
281+
## Huawei Cloud references
282+
283+
- [Creating an AOM alarm notification rule](https://support.huaweicloud.com/intl/en-us/usermanual-aom2/mon_01_0015.html)
284+
- [Configuring message templates](https://support.huaweicloud.com/intl/en-us/bestpractice-aom2/aom_05_0023.html)
285+
- [Subscribing to AOM metrics or alarms](https://support.huaweicloud.com/intl/en-us/usermanual-aom2/mon_01_0039.html)
286+
- [Pushing event data](https://support.huaweicloud.com/intl/en-us/api-aom/PushEvents.html)

‎integration-docs/src/doc-map.mjs‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -45,6 +45,7 @@ export const docMap = {
4545
Guance: `${alertBase}/guance.mdx`,
4646
Zilliz: `${alertBase}/zilliz.mdx`,
4747
HuaweiyunAPM: `${alertBase}/huawei-apm.mdx`,
48+
HuaweiyunAOM: `${alertBase}/huawei-aom.mdx`,
4849
ZStack: `${alertBase}/zstack.mdx`,
4950
Keep: `${alertBase}/keep.mdx`,
5051
ElastAlert2: `${alertBase}/elastalert2.mdx`,

0 commit comments

Comments
 (0)