Observability Audit

Feature Overview

Item	Content
Applicable Role	Operator
Navigation path	Settings > API Rate Control > Observability Audit
Page route	/user/system/rate-control/observation
Managed objects	Minute statistics, blocked logs, and audit logs

Beginner Explanation

Observability Audit works like a detail viewer for API rate control. Use it to check which requests were counted or blocked, then locate issues by rule, node, API, and time range.

Terms Quick Reference

Term	Description
Audit record	Details generated by rate-control statistics or blocked requests.; Filter by time during troubleshooting.
Node	The service node that executes rate-control rules.; Check node cache when node results differ.
Rule hit	A request matches and triggers a rule.; Compare with Rule Management.
Clean data	An action that deletes or archives audit data.; Confirm the scope before continuing.

Prerequisites

1. The current account has permission to access API rate-control observability audit data.
2. You have opened API Rate Control > Observability Audit.
3. The scope and purpose are confirmed before cleanup or export.

Page Description

The page provides API, node, and audit-log views for tracing rate-control decisions and abnormal requests.

Area	Description
Minute Statistics	View request, blocked, and over-limit statistics aggregated by minute.
Blocked Logs	View request records blocked by rate-control rules.
Audit Logs	View rule audit records.
API Path	Filter by API path.
Node ID	Filter by node.
Start Time / End Time	Filter by time range.
Clean	Entry for cleaning observability or audit data.
Export Statistics / Export Blocked Logs / Export Audit Logs	Export the corresponding data.

Main Operations

View Observability Metrics

1. Go to Settings > API Rate Control > Observability Audit and select a time range.
2. Filter by rule, node, request result, block reason, or request ID.
3. Check request volume, blocked requests, matched rules, and status distribution. If no data is shown, check the time zone, node, and filter scope.
4. Observability data may contain call identifiers and business paths and must be redacted before export or screenshots.

View Block Event Details

1. Click an abnormal metric or "Details" for the target event.
2. Review the matched rule, node, action, status code, time, and error summary.
3. Compare the rule version and conditions in Rule Management to determine whether the block was expected.
4. If it cannot be explained, record a redacted request ID, rule ID, and time range. Do not change rules or republish.

Use this operation to query observability and audit data. Do not add create, publish, or save operations to this query-oriented workflow.

Export Observability Data

1. Open Settings > API Rate Control > Observability Audit.
2. Locate the target Observability Audit and click "Export".
3. Review or complete the required fields shown on the page, and confirm the target object, scope, and current status.
4. For an action that changes data, permissions, status, or an external setting, confirm the impact and rollback path before clicking the final confirmation button.
5. After the action, return to the list or details page and verify the status, update time, or result message.

Clean Observability Data

1. Open Settings > API Rate Control > Observability Audit.
2. Locate the target Observability Audit and click "Clean".
3. Review or complete the required fields shown on the page, and confirm the target object, scope, and current status.
4. For an action that changes data, permissions, status, or an external setting, confirm the impact and rollback path before clicking the final confirmation button.
5. After the action, return to the list or details page and verify the status, update time, or result message.

Parameter Quick Reference

Field Name	Required	Field Type	Example	Description
Tab	Yes	Tab	Minute Statistics	Switches between statistics, blocked logs, and audit data views.
Time Range	No	Time range	Start Time / End Time	Filters observability or audit records by time.
API	No	Text	<api_path>	Filters by API path.
Node	No	Text	<node_name>	Filters by execution node.
Rule	No	Text	<rule_name>	Filters by matched rule.
Hit Count	System generated	Number	10	Shows rule or API hit count.
Blocked Result	System generated	Enum	Blocked	Shows whether the request was blocked by a rule.
Cost	System generated	Number	120 ms	Shows request or processing cost.
Audit Details	System generated	Text / link	View Details	Opens context information for an audit record.
Export	No	Button	Export Audit Logs	Exports statistics, blocked logs, or audit data.
Clean	No	Button	Clean	Cleans observability or audit data.

Pitfalls

• Observability audit data may expose API paths, nodes, rules, account clues, IP addresses, abnormal requests, and internal capacity information.
• Export Statistics, Export Blocked Logs, and Export Audit Logs export real audit data and are high-risk actions.
• Clean affects later troubleshooting and compliance retention and is a high-risk action.
• Do not write real API paths, tokens, IP addresses, accounts, tenant IDs, customer names, node names, internal error details, or load-test parameters in the manual.

Result Validation

Check Item	Success Signal	If Abnormal
Tab switching	The three data tabs can be switched normally.	Refresh the page and re-enter it.
Filter result	The list refreshes according to API, node, rule, or time range.	Check filters, click "Reset", and search again.
Export entry	Export buttons are displayed according to permissions.	Confirm time range, desensitization requirements, and recipient before exporting.
Clean entry	The cleanup entry is displayed according to permissions and has a clear scope confirmation before execution.	Stop cleanup if retention requirements are unclear.

FAQ

Overview shows blocked requests, but audit records are not found

The overview shows an increase in blocked requests, but no corresponding record appears in Observability Audit.

Possible cause:

The time ranges differ, or the current tab is not Blocked Logs.

Resolution:

Switch to Blocked Logs, expand the time range, and search again.

What should be checked before cleaning data?

The page provides a Clean entry.

Possible cause:

Cleanup affects later audit tracing.

Resolution:

Confirm retention requirements and troubleshooting needs first. Only an authorized operator should perform cleanup.

Why are observability audit records empty?

No API hit, rate-control, blocked-request, or exception record appears on the Observability Audit page.

Possible cause:

No request triggered a rate-control rule in the selected time range, requests bypassed the rate-control gateway, or audit collection is delayed.

Resolution:

Expand the time range and confirm the request path. Check rule publishing and gateway access. If records remain empty, compare reporting status in Node Cache.

How should the Observability Audit page be exported or captured safely?

Symptom:

Page information is needed for troubleshooting, audit, or delivery.

Possible causes:

The page may contain accounts, email addresses, IP addresses, internal paths, tenant identifiers, Keys, or amounts.

Resolution:

Keep only the necessary fields and action context. Use opaque light-gray pixel mosaics for sensitive text and never share complete credentials or internal addresses.

What should I do when the Observability Audit page shows unexpected data?

Symptom:

A field, status, metric, or related object differs from the expectation.

Possible causes:

The page scope, time condition, role permission, or upstream setting does not match.

Resolution:

Record the redacted object, time, and result. Verify the entry and filters first, then check related pages and Operation Logs.

Notes

• Observability audit data may expose API paths, nodes, rules, account clues, IP addresses, abnormal requests, and internal capacity information.
• Export Statistics, Export Blocked Logs, and Export Audit Logs export real audit data and are high-risk actions.
• Clean affects later troubleshooting and compliance retention and is a high-risk action.
• Do not write real API paths, tokens, IP addresses, accounts, tenant IDs, customer names, node names, internal error details, or load-test parameters in the manual.

Next Steps

1. To adjust rules, go to Rule Management.
2. To check node status, go to Node Cache.