Monitoring Multiple Instances of the File System

Describes how to monitor the health and performance of your cluster.

Determining the Number of Running Instances

Procedure

Run the following command to determine the number of instances actually running:
/opt/mapr/server/mrconfig info instances

Your output will look similar to the following. This output shows that two File Server instances are running on ports 5660 and 5661.

/opt/mapr/server/mrconfig info instances
    2
    5660 5661
Alternatively, on large clusters, run the following command to:
  • Determine the number of configured instances:
    maprcli node list -columns numInstances
        hostname           numInstances  ip            
        atsqa4-161.qa.lab  1             10.10.88.161  
        atsqa4-162.qa.lab  1             10.10.88.162  
        atsqa4-163.qa.lab  1             10.10.88.163  
        atsqa4-164.qa.lab  1             10.10.88.164
  • Determine the number of running instances reported by file system to CLDB:
    maprcli node list -columns numReportedInstances
        numReportedInstances  hostname           ip            
        2                     atsqa4-161.qa.lab  10.10.88.161  
        1                     atsqa4-162.qa.lab  10.10.88.162  
        1                     atsqa4-163.qa.lab  10.10.88.163  
        2                     atsqa4-164.qa.lab  10.10.88.164

Determining the Number of file system Threads

About this task

You can run the mrconfig info threads command to view file system threads from all the instances. The output is tagged to identify the instance.