Observability data storage
Application Real-Time Monitoring Service (ARMS) stores traces, metrics, logs, and profiling data in two backends: Simple Log Service (SLS) and Managed Service for Prometheus. Query the stored data in SLS or Prometheus to build dashboards, filter by dimensions, or aggregate across services.
Storage in SLS or Managed Service for Prometheus does not incur additional fees. You are charged only for using ARMS. For details, see Billing (new version).
Retention summary
The following table summarizes retention periods across all ARMS sub-services.
Sub-service | Data type | Retention |
Application Monitoring / Managed Service for OpenTelemetry | Metric data | 90 days |
Application Monitoring / Managed Service for OpenTelemetry | Trace data | 30 days |
Application Monitoring / Managed Service for OpenTelemetry | Profiling data | 7 days |
Real User Monitoring (RUM) | Log data | 30 days |
Real User Monitoring (RUM) | Alert metric data | 90 days |
Browser Monitoring | Observability data | 30 days |
Storage paths by sub-service
Each sub-service writes data to dedicated SLS LogProjects, Logstores, or Prometheus instances. Use the schema references to understand the field definitions for custom queries.
Do not delete the SLS LogProject or Prometheus instance that stores observability data. Deletion causes data loss and features become unavailable.
Application Monitoring and Managed Service for OpenTelemetry
Data type | Storage path | Schema reference |
Metric data |
| |
Trace data | SLS LogProject: | |
Profiling data | Custom development is not supported | -- |
Real User Monitoring (RUM)
Data type | Storage path | Schema reference |
Log data | SLS LogProject: | |
Alert metric data | Prometheus instance: |
Browser Monitoring
Data type | Storage path | Schema reference |
Observability data | SLS LogProject: |
The following describes the observability data processing and data storage resources for Application Monitoring and Real User Monitoring in CloudMonitor 2.0:
Computing resources
Cloud product
Task type
Purpose
Storage path
Application Monitoring/
Managed Service for OpenTelemetry
ScheduleSQL
Observability data processing
SLS Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
Schedule Job ID: A random string of up to 32 characters. Use the ID to view the description and script (SQL statement).
Real User Monitoring
ScheduleSQL
Observability data processing
SLS Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
Schedule Job ID: A random string of up to 32 characters. Use the ID to view the description and script (SQL statement).
Storage resources
Cloud product
Data type
Retention period
Storage path
Application Monitoring/
Managed Service for OpenTelemetry
Metrics data
90 days
SLS Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
SLS MetricStore:
metricstore-apm-metrics-detail: Storage medium for detailed metrics.
The name of the default attached Prometheus instance is metricstore-apm-metrics-detail_{region-id}_{workspace}
metricstore-apm-metrics: Storage medium for aggregated metrics.
The name of the default attached Prometheus instance is metricstore-apm-metrics_{region-id}_{workspace}
metricstore-apm-metrics-custom: Storage medium for business trace analysis and custom metrics.
The name of the default attached Prometheus instance is metricstore-apm-metrics-custom_{region-id}_{workspace}
Trace data
30 days
SLS Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
SLS Logstore: logstore-tracing
Profiling data
7 days
SLS Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
SLS Logstore: logstore-profiling
Real User Monitoring
Metrics data
90 days
SLS Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
SLS Metricstore: metricstore-rum-metrics
The name of the default attached Prometheus instance is rum-<region-id>
Log data
30 days
Project: proj-xtrace-<encode>-<region-id>
Example: proj-xtrace-a91c22e2d3d346a756da6faee7c450bb-cn-hangzhou
Logstore: logstore-rum
Storage path examples
The following examples show resolved storage paths for the China (Hangzhou) region:
Data type | Example path |
Application metrics |
|
Trace data (LogProject) |
|
RUM log data (LogProject) |
|
RUM alert metrics |
|
Browser Monitoring (LogProject) |
|
Browser Monitoring (Logstore) |
|
Query stored data
Use the schema references in the tables above to build SLS or Prometheus queries on your stored data:
Trace analysis: Query
logstore-tracingto investigate spans. See Trace Explorer parameters.Metric dashboards: Query the
arms_metrics_*Prometheus instance to build custom dashboards. See Application monitoring metrics.RUM log queries: Query
logstore-rumto analyze RUM log data. See Log data description.
Storage redundancy
ARMS Application Monitoring and RUM offer two types of data storage: locally redundant storage and zone-redundant storage. These storage types use a data redundancy mechanism for both single-zone and multi-zone deployments to ensure data durability and availability. ARMS is currently upgrading its storage redundancy feature. The availability of this feature varies by region, and its upgrade status is consistent with that of SLS. For region-specific details, see Storage redundancy.
Locally redundant storage and zone-redundant storage are offered with the same service availability and price on the ARMS data plane.
Storage redundancy type | Service availability |
Locally redundant storage | 99.9% |
Zone-redundant Storage | 99.9% |