Performance monitoring for CPFS for Lingjun - Cloud Parallel File Storage

The performance monitoring feature allows you to view real-time performance metrics for a file system or dataflow, such as read/write IOPS and throughput. This topic describes how to view this data for a CPFS for Lingjun file system.

Prerequisites

A file system must exist. For more information, see Create a file system.
You have activated Cloud Monitor.

You can go to the Cloud Monitor product page and follow the on-screen instructions to activate the service.

Usage notes

Client performance data can be viewed only in the Cloud Monitor console or by calling Cloud Monitor APIs. For more information, see View data in the Cloud Monitor console or Use Cloud Monitor APIs.
When you use a CPFS for Lingjun file system with Elastic Compute Service (ECS) or single-tenant PAI Lingjun resources, hostname refers to the hostname of the node.
When you use a CPFS for Lingjun file system with general-purpose Platform as a Service for AI (PAI) computing resources or PAI Lingjun resources, hostname refers to the pod ID of the task.

NAS console

Log on to the NAS console.
In the left-side navigation pane, choose Monitoring Audit > Performance Monitoring.
In the top navigation bar, select a region.
On the Performance Monitoring page, do the following:
1. From the File System Type drop-down list, select CPFS for Lingjun.
2. Select the ID of the file system from the File System ID drop-down list.
3. Select a time range, such as Hour, Day, 7 Days, 14 Days, Current Month, or Last Month, or specify a custom time range. The maximum time range is 30 days.
4. Click the All or Data Flow tab to view the performance of the file system or the dataflow's import and export tasks.

Cloud Monitor console

Log on to the Cloud Monitor console.
In the left-side navigation pane, choose .
On the Cloud Service Monitoring page, enter CPFS in the search box and click CPFS.
On the CPFS page, select a region, find the target file system, and then click Monitoring Charts in the Actions column.
On the Monitoring Charts page, click the File System Monitoring tab to view performance details of the target file system or the client.

Cloud Monitor API

You can also query monitoring data for CPFS by calling Cloud Monitor APIs. The primary APIs include:

DescribeMetricMetaList: queries the descriptions of time-series metrics available in Cloud Monitor.
DescribeMetricList: queries monitoring data for a cloud service's time-series metric over a specified period.
DescribeMetricLast: queries the latest monitoring data for a specified object.

The following table describes the request parameters for CPFS.

Parameter	Description
Namespace	The data namespace for the cloud service. For CPFS, the value is acs_nas.
MetricName	The name of the metric. Valid values: File system IopsRead IopsWrite LatencyRead LatencyWrite QpsMeta ThruputRead ThruputWrite Dataflow ThroughputImport ThroughputExport QPSImportMeta QPSExportMeta IOPSImport IOPSExport LatencyImport LatencyExport Client ClientReadIops ClientWriteIops ClientReadLatency ClientWriteLatency ClientReadThroughput ClientWriteThroughput
Dimensions	A map of dimensions used to query monitoring data for a specific resource. The format is as follows: `{"userId":"xxxxxx","fileSystemId":"bmcpfs-xxxxx"}`. Note The Dimensions parameter must be a JSON string representing the map object, and the dimensions must be in a specific order.

Related operations

You can configure an alert rule to receive timely notifications about abnormal data. This allows technical staff to respond and resolve issues promptly to ensure business continuity. For more information, see Configure a basic alert rule.
You can also create an application group to manage multiple file systems. This allows you to manage alert rules and view monitoring data at the group level, which simplifies management and improves monitoring efficiency. For more information, see Create an application group.