How to Monitor (Red Hat) Ceph Storage Systems with Storage Analyzer KM

Introduction

This article explains how to configure the Ceph (REST) adapter to monitor Red Hat Ceph storage systems using the REST API protocol.

Prerequisites

Before configuring the policy:

• Read these articles to make sure the prerequisites are met:

  • Storage Analyzer KM: Verifying Prerequisites (Credentials and Connectivity)
  • Required Roles and Permissions for Hardware and Storage Monitoring

• Run the ceph mgr services command on the Ceph system to determine the protocol/port that has been configured. In the example below, it’s https/8443:

  [root@ceph-storage2 ~]# ceph mgr services { “dashboard”: “https://ceph-storage2.internal.sentrysoftware.net:8443/” }

For optimal performance, we recommend reserving 2 CPU cores and at least 4 GB of RAM for operating the Storage Analyzer KM. The amount of memory allocated to the Java engine of the KM may be adjusted via the /SENTRY/STORAGE/collectionHubHeapSizeMax configuration variable. For more details, refer to:

• Advanced Configuration Variables
• Storage KMs Run Out-Of-Memory and Stop Collecting Data.

Procedure

1. Log in to BMC TrueSight/Helix Operations Management

2. Create a new monitoring policy:

   • Specify the general properties and agent selection criteria
   • Add a monitoring configuration
   • Select the Storage Analyzer solution
   • In the Storage Configuration section, click Add to create a new Endpoint

   

3. Configure the Endpoint Settings:

   • Enter an Internal ID
   • Specify the Hostname/IP address of the Ceph storage system
   • Select CEPH (REST) from the Storage Intelligence Adapter dropdown list
   • Untick the Hardware Monitoring checkbox
   • (Optional) Specify a custom device name in the Associate Device FQDN in TrueSight field. For more details, refer to More Easily Identify Your Devices Monitored Through a Proxy in TrueSight / Helix Operations Management.

   

4. Configure the Endpoint Connection Settings:

   • Provide the Username and Password
   • Specify the port (7000 or 8443 by default). See the output of ceph mgr services above
   • Check Encryption if using the HTTPS protocol. See the output of ceph mgr services above

   

5. Configure the Advanced Settings:

   • In large environments, we recommend increasing the values of the Discovery Interval and Polling Intervals to maintain the workload of the PATROL Agent and monitored systems at a reasonable level.
   • Since volume monitoring is highly resource-intensive, we recommend disabling it if not needed to reduce the workload on the KM and the PATROL Agent.
   • Because displaying thousands of volumes will have a huge impact on the PATROL Agent and on BMC TrueSight/Helix, volumes are discovered but NOT displayed. Uncheck the Hide Volume Instances box if required.

   

6. Save the policy and verify that it applies to the right PATROL Agent.

7. Allow the KM some time to discover your storage system(s). Access the Devices page and search for their hostname, IP address, FQDN, or custom MetaFQDN/Device Name. You can also refer to the article Identifying and Locating Devices in TrueSight to learn how to run the below PSL script in the Query PATROL Agent window:

       CLASS_NAME = "SKM_SYSTEM";
       nodes = get_vars("/".CLASS_NAME, "subnodes");
       foreach node (nodes)
       {
       objectPath = "/".CLASS_NAME."/".node;
       instanceName = get(objectPath."/MetaFQDN");
       output  = [output, "Device Name:  ".instanceName ];
       }
       print(output);
   

Related Topics

• How to Monitor Dell EMC Unity Storage Systems with Storage Analyzer KM
• How to Monitor Hitachi E and G Series Storage Systems with Storage Analyzer KM
• How to Monitor Pure Storage Systems with Storage Analyzer KM
• How to Monitor Dell EMC ScaleIO/PowerFlex Storage Systems with Storage Analyzer KM
• How to Monitor NetApp Storage Systems via NetApp Active IQ Unified Manager with Storage Analyzer KM