List of queries

You can make the following predefined queries with query option of the mmperfmon command.

General and network

  • usage: Retrieves details about the CPU, memory, storage, and network usage.
  • cpu: Retrieves details of the CPU usage in system and user space, and context switches.
  • netDetails: Retrieves details about the network.
  • netErrors: Retrieves details about network problems, such as collisions, drops, and errors, for all available networks.
  • compareNodes: Compares a single metric across all nodes that are running sensors.

GPFS

GPFS metric queries give an overall view of the GPFS without considering the protocols.
  • gpfsCRUDopsLatency: Retrieves information about the GPFS create, retrieve, update, and delete operations latency.
  • gpfsFSWaits: Retrieves information on the maximum waits for read and write operations for all file systems.
  • gpfsNSDWaits: Retrieves information on the maximum waits for read and write operations for all disks.
  • gpfsNumberOperations: Retrieves the number of operations to the GPFS file system.
  • gpfsVFSOpCounts: Retrieves VFS operation counts.

Cross protocol

These queries retrieve information after metrics are compared between different protocols on a particular node.
  • protocolIOLatency: Compares latency per protocol (SMB and NFS).
  • protocolIORate: Retrieves the percentage of total I/O rate per protocol (SMB and NFS,).
  • protocolThroughput: Retrieves the percentage of total throughput per protocol (SMB and NFS).

NFS

These queries retrieve metrics associated with the NFS protocol.
  • nfsIOLatency: Retrieves the NFS I/O latency in nanoseconds.
  • nfsIORate: Retrieves the NFS I/O operations per second (NFS IOPS).
  • nfsThroughput: Retrieves the NFS throughput in bytes per second.
  • nfsErrors: Retrieves the NFS error count for read and write operations.
  • nfsQueue: Retrieves the NFS read and write queue latency in nanoseconds.
  • nfsThroughputPerOp: Retrieves the NFS read and write throughput per operation in bytes.

SMB

These queries retrieve metrics associated with SMB.
  • smb2IOLatency: Retrieves the SMB2 I/O latencies per bucket size (default 1 sec).
  • smb2IORate: Retrieves the SMB2 I/O rate in number of operations per bucket size (default 1 sec).
  • smb2Throughput: Retrieves the SMB2 Throughput in bytes per bucket size (default 1 sec).
  • smb2Writes: Retrieves count, # of idle calls, bytes in and out, and operation time for SMB2 writes.
  • smbConnections: Retrieves the number of SMB connections.

CTDB

These queries retrieve metrics associated with CTDB.
  • ctdbCallLatency: Retrieves information on the CTDB call latency.
  • ctdbHopCountDetails: Retrieves information on the CTDB hop count buckets 0 - 5 for one database.
  • ctdbHopCounts: Retrieves information on the CTDB hop counts (bucket 00 = 1 - 3 hops) for all databases.