GPFS metrics

The following section lists all the GPFS metrics:

GPFSDisk

gpfs_ds_bytes_read: Number of bytes read.
gpfs_ds_bytes_written: Number of bytes written.
gpfs_ds_max_disk_wait_rd: The longest time spent waiting for a disk read operation.
gpfs_ds_max_disk_wait_wr: The longest time spent waiting for a disk write operation.
gpfs_ds_max_queue_wait_rd: The longest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_ds_max_queue_wait_wr: The longest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_ds_min_disk_wait_rd: The shortest time spent waiting for a disk read operation.
gpfs_ds_min_disk_wait_wr: The shortest time spent waiting for a disk write operation.
gpfs_ds_min_queue_wait_rd: The shortest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_ds_min_queue_wait_wr: The shortest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_ds_read_ops: Number of read operations.
gpfs_ds_tot_disk_wait_rd: The total time in seconds spent waiting for disk read operations.
gpfs_ds_tot_disk_wait_wr: The total time in seconds spent waiting for disk write operations.
gpfs_ds_tot_queue_wait_rd: The total time that is spent between being enqueued for a read operation and the completion of that operation.
gpfs_ds_tot_queue_wait_wr: The total time that is spent between being enqueued for a write operation and the completion of that operation.
gpfs_ds_write_ops: Number of write operations.

GPFSDiskCap

Specifies the available disk space capacity on GPFS file systems per pool and per disk.

The key structure is:

Following are the metrics (metric_name):

gpfs_disk_disksize: Total size of disk.
gpfs_disk_free_fullkb: Total available space on disk in full blocks.
gpfs_disk_free_fragkb: Total available space on disk in fragments.

GPFSInodeCap

Specifies the available inode capacity on GPFS file systems.

The key structure is:

<gpfs_cluster_name>|GPFSInodeCap|<gpfs_fs_name>|<metric_name>

Following are the metrics (metric_name):

gpfs_fs_inode_used: Number of used inodes.
gpfs_fs_inode_free: Number of free inodes.
gpfs_fs_inode_alloc: Number of allocated inodes.
gpfs_fs_inode_max: Maximum number of inodes.

Note: The GPFSInodeCap is not an independent sensor in the perfmon config, but a sub-sensor of the GPFSDiskCap sensor.

GPFSPoolCap

Specifies the available disk space capacity on GPFS file systems per pool and per disk usage type.

The key structure is:

The gpfs_disk_usage_name can be either of the following values:

dataAndMetadata
dataOnly
descOnly
metadataOnly

Following are the metrics (metric_name):

gpfs_pool_disksize: Total size of all disks for this usage type.
gpfs_pool_free_fullkb: Total available disk space in full blocks for this usage type.
gpfs_pool_free_fragkb: Total available space in fragments for this usage type.

Note: The GPFSPoolCap is not an independent sensor in the perfmon config, but a sub-sensor of the GPFSDiskCap sensor.

GPFSFileset

For each independent fileset in the file system: Cluster name - GPFSFileset - filesystem name - fileset name.
For example, myCluster|GPFSFileset|myFilesystem|myFileset|gpfs_fset_maxInodes.

gpfs_fset_maxInodes: Maximum number of inodes for this independent fileset.
gpfs_fset_freeInodes: Number of free inodes available for this independent fileset.
gpfs_fset_allocInodes: Number of inodes allocated for this independent fileset.

GPFSFileSystem

For each file system, for example myMachine|GPFSFilesystem|myCluster|myFilesystem|gpfs_fs_bytes_read

gpfs_fs_bytes_read: Number of bytes read.
gpfs_fs_bytes_written: Number of bytes written.
gpfs_fs_disks: Number of disks in the file system.
gpfs_fs_max_disk_wait_rd: The longest time spent waiting for a disk read operation.
gpfs_fs_max_disk_wait_wr: The longest time spent waiting for a disk write operation.
gpfs_fs_max_queue_wait_rd: The longest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_fs_max_queue_wait_wr: The longest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_fs_min_disk_wait_rd: The shortest time spent waiting for a disk read operation.
gpfs_fs_min_disk_wait_wr: The shortest time spent waiting for a disk write operation.
gpfs_fs_min_queue_wait_rd: The shortest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_fs_min_queue_wait_wr: The shortest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_fs_read_ops: Number of read operations.
gpfs_fs_tot_disk_wait_rd: The total time in seconds spent waiting for disk read operations.
gpfs_fs_tot_disk_wait_wr: The total time in seconds spent waiting for disk write operations.
gpfs_fs_tot_queue_wait_rd: The total time that is spent between being enqueued for a read operation and the completion of that operation.
gpfs_fs_tot_queue_wait_wr: The total time that is spent between being enqueued for a write operation and the completion of that operation.
gpfs_fs_write_ops: Number of write operations.

Note:

The behavior of the minimum and maximum wait time for read and write I/O to disk and the queue wait time, for example metrics such as *max_disk_wait_wr and *max_queue_wait_wr, has changed. These metrics are now reset each time a sample is taken.

In the previous releases, these metrics were the minimum or maximum values noted since the start of the mmfsd daemon and were only reset after the mmfsd daemon was restarted. The maximum and minimum values are equal to the maximum and minimum value of each read instance in a sensor period because the metrics are reset each time a sample is taken in a sensor period.

GPFSFileSystemAPI

These metrics give the following information for each file system (application view). For example, myMachine|GPFSFilesystemAPI|myCluster|myFilesystem|gpfs_fis_bytes_read.

gpfs_fis_bytes_read: Number of bytes read.
gpfs_fis_bytes_written: Number of bytes written.
gpfs_fis_close_calls: Number of close calls.
gpfs_fis_disks: Number of disks in the file system.
gpfs_fis_inodes_written: Number of inode updates to disk.
gpfs_fis_open_calls: Number of open calls.
gpfs_fis_read_calls: Number of read calls.
gpfs_fis_readdir_calls: Number of readdir calls.
gpfs_fis_write_calls: Number of write calls.

GPFSLROC

These metrics provide following details about the Local Read Only Cache (LROC) operations:

For example, myNode|GPFSLROC|gpfs_lroc_capmb.

Note: All minimum (*_rmin), maximum (*_rmax), and average (*_ravg) result time metrics can be reset only once when the mmfsd daemon is restarted, or using the mmpmon reset or mmfsadm resetstats commands.

gpfs_lroc_agt_i: Total number of inserts from the FLEA agent.
gpfs_lroc_agt_i_ravg: Average response time in microseconds, which is taken by FLEA agent to insert the data.
gpfs_lroc_agt_i_rmax: Maximum response time in microseconds, which is taken by FLEA agent to insert the data.
gpfs_lroc_agt_i_rmin: Minimum response time in microseconds, which is taken by FLEA agent to insert the data.
gpfs_lroc_agt_r: Total number of reads from the FLEA agent.
gpfs_lroc_agt_r_ravg: Average response time in microseconds, which is taken by FLEA agent to read the data.
gpfs_lroc_agt_r_rmax: Maximum response time in microseconds, which is taken by FLEA agent to read the data.
gpfs_lroc_agt_r_rmin: Minimum response time in microseconds, which is taken by FLEA agent to read the data.
gpfs_lroc_capmb: Total capacity of space in MB.
gpfs_lroc_data_abs: Average amount of data in bytes that is stored in the cache per buffer descriptor.
gpfs_lroc_data_i: Total number of data invalidations in the cache.
gpfs_lroc_data_if: Total number of failures to invalidate data from the cache.
gpfs_lroc_data_imb: Total amount of invalidated data in MB.
gpfs_lroc_data_inr: Total number of data invalidations without being recalled from the cache.
gpfs_lroc_data_q: Total number of data queries in the cache.
gpfs_lroc_data_qf: Total number of data query failures.
gpfs_lroc_data_qmb: Total amount of data in MB, which is queried in the cache.
gpfs_lroc_data_r: Total number of data that is recalled from the cache.
gpfs_lroc_data_rf: Total number of failures that are required to recall the data.
gpfs_lroc_data_rmb: Total amount of data in MB, which is recalled from the cache.
gpfs_lroc_data_s: Total number of data that is stored in the cache.
gpfs_lroc_data_sf: Total number of failures to store data in the cache.
gpfs_lroc_data_smb: Total amount of space in MB used to cache the data.
gpfs_lroc_directory_i: Total number of directories that are invalidated from the cache.
gpfs_lroc_directory_if: Total number of failures that is required to invalidate directory data from the cache.
gpfs_lroc_directory_imb: Total amount of invalidated directory data in MB.
gpfs_lroc_directory_inr: Total number of directories that are invalidated without being recalled from the cache.
gpfs_lroc_directory_q: Total number of directories that are queried in the cache.
gpfs_lroc_directory_qf: Total number of directory data query failures.
gpfs_lroc_directory_qmb: Total amount of directory data in MB, which is queried in the cache.
gpfs_lroc_directory_r: Total number of directories that are recalled from the cache.
gpfs_lroc_directory_rf: Total number of failures to recall the directory data.
gpfs_lroc_directory_rmb: Total amount of directory data in MB, which is recalled from the cache.
gpfs_lroc_directory_s: Total number of directories that are stored in the cache.
gpfs_lroc_directory_sf: Total number of failures to store the directory data in the cache.
gpfs_lroc_directory_smb: Total amount of space in MB, which is used to cache the directory data.
gpfs_lroc_inode_i: Total number of inodes that are invalidated from the cache.
gpfs_lroc_inode_if: Total number of failures to invalidate the inode data from the cache.
gpfs_lroc_inode_imb: Total amount of space in MB to invalidate the inode data from the cache.
gpfs_lroc_inode_inr: Total number of inodes that are invalidated without being recalled from the cache.
gpfs_lroc_inode_q: Total number of inodes that are queried in the cache.
gpfs_lroc_inode_qf: Total number of inode query failures.
gpfs_lroc_inode_qmb: Total amount of space in MB of inode data that is queried in the cache.
gpfs_lroc_inode_r: Total number of inodes that are recalled from the cache.
gpfs_lroc_inode_rf: Total number of failures to recall inode data.
gpfs_lroc_inode_rmb: Total amount of space in MB of inode data that is recalled from the cache.
gpfs_lroc_inode_s: Total number of inodes that are stored in the cache.
gpfs_lroc_inode_sf: Total number of failures to store inode data in the cache.
gpfs_lroc_inode_smb: Total amount of space in MB that is used to cache the inodes.
gpfs_lroc_ssd_r: Total number of IOPs that are read into the SSD device.
gpfs_lroc_ssd_r_ravg: Average response time in microseconds that is taken by the SSD device to read the data.
gpfs_lroc_ssd_r_rmax: Maximum response time in microseconds, which is taken by the SSD device to read the data.
gpfs_lroc_ssd_r_rmin: Minimum response time in microseconds, which is taken by the SSD device to read the data.
gpfs_lroc_ssd_r_p: Total number of pages that are read from the SSD device.
gpfs_lroc_ssd_w: Total number of IOPs that are written to the SSD device.
gpfs_lroc_ssd_w_p: Total number of pages that are written to the SSD device.
gpfs_lroc_ssd_w_ravg: Average response time in microseconds, which is taken by the SSD device to write the data.
gpfs_lroc_ssd_w_rmax: Maximum response time in microseconds, which is taken by the SSD device to write the data.
gpfs_lroc_ssd_w_rmin: Minimum response time in microseconds, which is taken by the SSD device to write the data.
gpfs_lroc_usemb: Total used space in MB.

GPFSNode

These metrics give the following information for a particular node. For example, myNode|GPFSNode|gpfs_ns_bytes_read.

gpfs_ns_bytes_read: Number of bytes read.
gpfs_ns_bytes_written: Number of bytes written.
gpfs_ns_clusters: Number of clusters that are participating.
gpfs_ns_disks: Number of disks in all mounted file systems.
gpfs_ns_filesys: Number of mounted file systems.
gpfs_ns_max_disk_wait_rd: The longest time spent waiting for a disk read operation.
gpfs_ns_max_disk_wait_wr: The longest time spent waiting for a disk write operation.
gpfs_ns_max_queue_wait_rd: The longest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_ns_max_queue_wait_wr: The longest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_ns_min_disk_wait_rd: The shortest time spent waiting for a disk read operation.
gpfs_ns_min_disk_wait_wr: The shortest time spent waiting for a disk write operation.
gpfs_ns_min_queue_wait_rd: The shortest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_ns_min_queue_wait_wr: The shortest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_ns_read_ops: Number of read operations.
gpfs_ns_tot_disk_wait_rd: The total time in seconds spent waiting for disk read operations.
gpfs_ns_tot_disk_wait_wr: The total time in seconds spent waiting for disk write operations.
gpfs_ns_tot_queue_wait_rd: The total time that is spent between being enqueued for a read operation and the completion of that operation.
gpfs_ns_tot_queue_wait_wr: The total time that is spent between being enqueued for a write operation and the completion of that operation.
gpfs_ns_write_ops: Number of write operations.

Note:

GPFSNodeAPI

These metrics give the following information for a particular node from its application point of view. For example, myMachine|GPFSNodeAPI|gpfs_is_bytes_read.

gpfs_is_bytes_read: Number of bytes read.
gpfs_is_bytes_written: Number of bytes written.
gpfs_is_close_calls: Number of close calls.
gpfs_is_inodes_written: Number of inode updates to disk.
gpfs_is_open_calls: Number of open calls.
gpfs_is_readDir_calls: Number of readdir calls.
gpfs_is_read_calls: Number of read calls.
gpfs_is_write_calls: Number of write calls.

GPFSNSDDisk

These metrics give the following information about each NSD disk on the NSD server. For example, myMachine|GPFSNSDDisk|myNSDDisk|gpfs_nsdds_bytes_read.

gpfs_nsdds_bytes_read: Number of bytes read.
gpfs_nsdds_bytes_written: Number of bytes written.
gpfs_nsdds_max_disk_wait_rd: The longest time spent waiting for a disk read operation.
gpfs_nsdds_max_disk_wait_wr: The longest time spent waiting for a disk write operation.
gpfs_nsdds_max_queue_wait_rd: The longest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_nsdds_max_queue_wait_wr: The longest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_nsdds_min_disk_wait_rd: The shortest time spent waiting for a disk read operation.
gpfs_nsdds_min_disk_wait_wr: The shortest time spent waiting for a disk write operation.
gpfs_nsdds_min_queue_wait_rd: The shortest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_nsdds_min_queue_wait_wr: The shortest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_nsdds_read_ops: Number of read operations.
gpfs_nsdds_tot_disk_wait_rd: The total time in seconds spent waiting for disk read operations.
gpfs_nsdds_tot_disk_wait_wr: The total time in seconds spent waiting for disk write operations.
gpfs_nsdds_tot_queue_wait_rd: The total time that is spent between being enqueued for a read operation and the completion of that operation.
gpfs_nsdds_tot_queue_wait_wr: The total time that is spent between being enqueued for a write operation and the completion of that operation.
gpfs_nsdds_write_ops: Number of write operations.

Note:

GPFSNSDFS

These metrics give the following information for each file system served by a specific NSD server. For example, myMachine|GPFSNSDFS|myFilesystem|gpfs_nsdfs_bytes_read.

gpfs_nsdfs_bytes_read: Numbers of NSD bytes read, aggregated to the file system.
gpfs_nsdfs_bytes_written: Numbers of NSD bytes written, aggregated to the file system.
gpfs_nsdfs_read_ops: Numbers of NSD read operations, aggregated to the file system.
gpfs_nsdfs_write_ops: Numbers of NSD write operations, aggregated to the file system.

GPFSNSDPool

These metrics give the following information for each file system and pool that is served by a specific NSD server. For example, myMachine|GPFSNSDPool|myFilesystem|myPool|gpfs_nsdpool_bytes_read.

gpfs_nsdpool_bytes_read: Numbers of NSD bytes read, aggregated to the file system.
gpfs_nsdpool_bytes_written: Numbers of NSD bytes written, aggregated to the file system.
gpfs_nsdpool_read_ops: Numbers of NSD read operations, aggregated to the file system.
gpfs_nsdpool_write_ops: Numbers of NSD write operations, aggregated to the file system.

GPFSPDDisk

The sensor GPFSPDDisk provides Pdisk server statistics.

gpfs_pdds_read_ops: The number of disk read operations.
gpfs_pdds_bytes_read: The number of bytes read from disk.
gpfs_pdds_tot_disk_wait_rd: The total time spent waiting for disk read operations.
gpfs_pdds_tot_queue_wait_rd: The total time spent between being enqueued for a disk read operation and the completion of that operation.
gpfs_pdds_min_disk_wait_rd: The shortest time spent waiting for a disk read operation.
gpfs_pdds_min_queue_wait_rd: The shortest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_pdds_max_disk_wait_rd: The longest time spent waiting for a disk read operation.
gpfs_pdds_max_queue_wait_rd: The longest time between being enqueued for a disk read operation and the completion of that operation.
gpfs_pdds_write_ops: The number of disk write operations
gpfs_pdds_bytes_written: The number of bytes written to disk
gpfs_pdds_tot_disk_wait_wr: The total time spent waiting for disk write operations.
gpfs_pdds_tot_queue_wait_wr: The total time spent between being enqueued for a disk write operation and the completion of that operation.
gpfs_pdds_min_disk_wait_wr: The shortest time spent waiting for a disk write operation.
gpfs_pdds_min_queue_wait_wr: The shortest time between being enqueued for a disk write operation and the completion of that operation.
gpfs_pdds_max_disk_wait_wr: The longest time spent waiting for a disk write operation.
gpfs_pdds_max_queue_wait_wr: The longest time between being enqueued for a disk write operation and the completion of that operation.

GPFSPool

For each pool in each file system: Cluster name - GPFSPool - filesystem name -pool name.
For example, myCluster|GPFSPool|myFilesystem|myPool|gpfs_pool_free_dataKBvalid*.

gpfs_pool_free_dataKB: Free capacity for data (in KB) in the pool.
gpfs_pool_total_dataKB: Total capacity for data (in KB) in the pool.
gpfs_pool_free_metaKB: Free capacity for metadata (in KB) in the pool.
gpfs_pool_total_metaKB: Total capacity for metadata (in KB) in the pool.

GPFSPoolIO

gpfs_pool_bytes_rd: Bytes read from the pool.
gpfs_pool_bytes_wr: Bytes written to the pool.
gpfs_pool_free_fragkb: Total available space in fragments for this usage type.

GPFSQoS

gpfs_qos_et: The interval in seconds during which the measurement was made.
gpfs_qos_iops: The performance of the class in I/O operations per second.
gpfs_qos_ioql: The average number of I/O requests in the class that are pending for reasons other than being queued by QoS.
gpfs_qos_maxiops:: The limitation of the class in I/O operations per second.
gpfs_qos_maxmbs: The limitation of the class in I/O throughput per second.
gpfs_qos_mbs: I/O in megabytes per second.
gpfs_qos_nnodes: The number of nodes of throttling of the class are widely opened.
gpfs_qos_qsdl: The average number of I/O requests in the class that are queued by QoS.

GPFSVFSX

The GPFSVFSX sensor provides virtual file system operations statistics, including the number of operations and their average, minimum and maximum latency metrics.

Some performance monitoring sensors, such as VFS, are not enabled by default, even though they have predefined queries that are associated with the mmperfmon query command. This happens because the collector might have performance issues when it is required to collect more than a million metrics per second.

In order to enable the VFS statistics and sensor, use the mmfsadm vfsstats enable command on the node. Similarly, to enable the GPFSVFSX sensor, set the period value to an integer n, which is greater than zero by using the following command:

mmperfmon config update GPFSVFSX.period=n

Later, the VFS statistics can be disabled by using the mmfsadm vfsstats disable command, and the GPFSVFSX sensor can be disabled by using the following command:

mmperfmon config update GPFSVFSX.period=0

The GPFSVFSX sensor provides the following metrics:

gpfs_vfsx_accesses: Number of accesses operations.
gpfs_vfsx_accesses_t: Amount of time in seconds spent in accesses operations.
gpfs_vfsx_accesses_tmax: Maximum amount of time in seconds spent in accesses operations.
gpfs_vfsx_accesses_tmin: Minimum amount of time in seconds spent in accesses operations.
gpfs_vfsx_aioread: Number of aioread operations.
gpfs_vfsx_aioread_t: Amount of time in seconds spent in aioread operations.
gpfs_vfsx_aioread_tmax: Maximum amount of time in seconds spent in aioread operations.
gpfs_vfsx_aioread_tmin: Minimum amount of time in seconds spent in aioread operations.
gpfs_vfsx_aiowrite: Number of aiowrite operations.
gpfs_vfsx_aiowrite_t: Amount of time in seconds spent in aiowrite operations.
gpfs_vfsx_aiowrite_tmax: Maximum amount of time in seconds spent in aiowrite operations.
gpfs_vfsx_aiowrite_tmin: Minimum amount of time in seconds spent in aiowrite operations.
gpfs_vfsx_clear: Number of clear operations.
gpfs_vfsx_clear_t: Amount of time in seconds spent in clear operations.
gpfs_vfsx_clear_tmax: Maximum amount of time in seconds spent in clear operations.
gpfs_vfsx_clear_tmin: Minimum amount of time in seconds spent in clear operations.
gpfs_vfsx_close: Number of close operations.
gpfs_vfsx_close_t: Amount of time in seconds spent in close operations.
gpfs_vfsx_close_tmax: Maximum amount of time in seconds spent in close operations.
gpfs_vfsx_close_tmin: Minimum amount of time in seconds spent in close operations.
gpfs_vfsx_create: Number of create operations.
gpfs_vfsx_create_t: Amount of time in seconds spent in create operations.
gpfs_vfsx_create_tmax: Maximum amount of time in seconds spent in create operations.
gpfs_vfsx_create_tmin: Minimum amount of time in seconds spent in create operations.
gpfs_vfsx_decodeFh: Number of decodeFh operations.
gpfs_vfsx_decodeFh_t: Amount of time in seconds spent in decodeFh operations.
gpfs_vfsx_decodeFh_tmax: Maximum amount of time in seconds spent in decodeFh operations.
gpfs_vfsx_decodeFh_tmin: Minimum amount of time in seconds spent in decodeFh operations.
gpfs_vfsx_encodeFh: Number of encodeFh operations.
gpfs_vfsx_encodeFh_t: Amount of time in seconds spent in encodeFh operations.
gpfs_vfsx_encodeFh_tmax: Maximum amount of time in seconds spent in encodeFh operations.
gpfs_vfsx_encodeFh_tmin: Minimum amount of time in seconds spent in encodeFh operations.
gpfs_vfsx_flock: Number of flock operations.
gpfs_vfsx_flock_t: Amount of time in seconds spent in flock operations.
gpfs_vfsx_flock_tmax: Maximum amount of time in seconds spent in flock operations.
gpfs_vfsx_flock_tmin: Minimum amount of time in seconds spent in flock operations.
gpfs_vfsx_fsync: Number of fsync operations.
gpfs_vfsx_fsync_t: Amount of time in seconds spent in fsync operations.
gpfs_vfsx_fsync_tmax: Maximum amount of time in seconds spent in fsync operations.
gpfs_vfsx_fsync_tmin: Minimum amount of time in seconds spent in fsync operations.
gpfs_vfsx_fsyncRange: Number of fsyncRange operations.
gpfs_vfsx_fsyncRange_t: Amount of time in seconds spent in fsyncRange operations.
gpfs_vfsx_fsyncRange_tmax: Maximum amount of time in seconds spent in fsyncRange operations.
gpfs_vfsx_fsyncRange_tmin: Minimum amount of time in seconds spent in fsyncRange operations.
gpfs_vfsx_ftrunc: Number of ftrunc operations.
gpfs_vfsx_ftrunc_t: Amount of time in seconds spent in ftrunc operations.
gpfs_vfsx_ftrunc_tmax: Maximum amount of time in seconds spent in ftrunc operations.
gpfs_vfsx_ftrunc_tmin: Minimum amount of time in seconds spent in ftrunc operations.
gpfs_vfsx_getattr: Number of getattr operations.
gpfs_vfsx_getattr_t: Amount of time in seconds spent in getattr operations.
gpfs_vfsx_getattr_tmax: Maximum amount of time in seconds spent in getattr operations.
gpfs_vfsx_getattr_tmin: Minimum amount of time in seconds spent in getattr operations.
gpfs_vfsx_getDentry: Number of getDentry operations.
gpfs_vfsx_getDentry_t: Amount of time in seconds spent in getDentry operations.
gpfs_vfsx_getDentry_tmax: Maximum amount of time in seconds spent in getDentry operations.
gpfs_vfsx_getDentry_tmin: Minimum amount of time in seconds spent in getDentry operations.
gpfs_vfsx_getParent: Number of getParent operations.
gpfs_vfsx_getParent_t: Amount of time in seconds spent in getParent operations.
gpfs_vfsx_getParent_tmax: Maximum amount of time in seconds spent in getParent operations.
gpfs_vfsx_getParent_tmin: Minimum amount of time in seconds spent in getParent operations.
gpfs_vfsx_getxattr: Number of getxattr operations.
gpfs_vfsx_getxattr_t: Amount of time in seconds spent in getxattr operations.
gpfs_vfsx_getxattr_tmax: Maximum amount of time in seconds spent in getxattr operations.
gpfs_vfsx_getxattr_tmin: Minimum amount of time in seconds spent in getxattr operations.
gpfs_vfsx_link: Number of link operations.
gpfs_vfsx_link_t: Amount of time in seconds spent in link operations.
gpfs_vfsx_link_tmax: Maximum amount of time in seconds spent in link operations.
gpfs_vfsx_link_tmin: Minimum amount of time in seconds spent in link operations.
gpfs_vfsx_listxattr: Number of listxattr operations.
gpfs_vfsx_listxattr_t: Amount of time in seconds spent in listxattr operations.
gpfs_vfsx_listxattr_tmax: Maximum amount of time in seconds spent in listxattr operations.
gpfs_vfsx_listxattr_tmin: Minimum amount of time in seconds spent in listxattr operations.
gpfs_vfsx_lockctl: Number of lockctl operations.
gpfs_vfsx_lockctl_t: Amount of time in seconds spent in lockctl operations.
gpfs_vfsx_lockctl_tmax: Maximum amount of time in seconds spent in lockctl operations.
gpfs_vfsx_lockctl_tmin: Minimum amount of time in seconds spent in lockctl operations.
gpfs_vfsx_lookup: Number of lookup operations.
gpfs_vfsx_lookup_t: Amount of time in seconds spent in lookup operations.
gpfs_vfsx_lookup_tmax: Maximum amount of time in seconds spent in lookup operations.
gpfs_vfsx_lookup_tmin: Minimum amount of time in seconds spent in lookup operations.
gpfs_vfsx_mapLloff: Number of mapLloff operations.
gpfs_vfsx_mapLloff_t: Amount of time in seconds spent in mapLloff operations.
gpfs_vfsx_mapLloff_tmax: Maximum amount of time in seconds spent in mapLloff operations.
gpfs_vfsx_mapLloff_tmin: Minimum amount of time in seconds spent in mapLloff operations.
gpfs_vfsx_mkdir: Number of mkdir operations.
gpfs_vfsx_mkdir_t: Amount of time in seconds spent in mkdir operations.
gpfs_vfsx_mkdir_tmax: Maximum amount of time in seconds spent in mkdir operations.
gpfs_vfsx_mkdir_tmin: Minimum amount of time in seconds spent in mkdir operations.
gpfs_vfsx_mknod: Number of mknod operations.
gpfs_vfsx_mknod_t: Amount of time in seconds spent in mknod operations.
gpfs_vfsx_mknod_tmax: Maximum amount of time in seconds spent in mknod operations.
gpfs_vfsx_mknod_tmin: Minimum amount of time in seconds spent in mknod operations.
gpfs_vfsx_mmapread: Number of mmapread operations.
gpfs_vfsx_mmapread_t: Amount of time in seconds spent in mmapread operations.
gpfs_vfsx_mmapread_tmax: Maximum amount of time in seconds spent in mmapread operations.
gpfs_vfsx_mmapread_tmin: Minimum amount of time in seconds spent in mmapread operations.
gpfs_vfsx_mmapwrite: Number of mmapwrite operations.
gpfs_vfsx_mmapwrite_t: Amount of time in seconds spent in mmapwrite operations.
gpfs_vfsx_mmapwrite_tmax: Maximum amount of time in seconds spent in mmapwrite operations.
gpfs_vfsx_mmapwrite_tmin: Minimum amount of time in seconds spent in mmapwrite operations.
gpfs_vfsx_mount: Number of mount operations.
gpfs_vfsx_mount_t: Amount of time in seconds spent in mount operations.
gpfs_vfsx_mount_tmax: Maximum amount of time in seconds spent in mount operations.
gpfs_vfsx_mount_tmin: Minimum amount of time in seconds spent in mount operations.
gpfs_vfsx_open: Number of open operations.
gpfs_vfsx_open_t: Amount of time in seconds spent in open operations.
gpfs_vfsx_open_tmax: Maximum amount of time in seconds spent in open operations.
gpfs_vfsx_open_tmin: Minimum amount of time in seconds spent in open operations.
gpfs_vfsx_read: Number of read operations.
gpfs_vfsx_read_t: Amount of time in seconds spent in read operations.
gpfs_vfsx_read_tmax: Maximum amount of time in seconds spent in read operations.
gpfs_vfsx_read_tmin: Minimum amount of time in seconds spent in read operations.
gpfs_vfsx_readdir: Number of readdir operations.
gpfs_vfsx_readdir_t: Amount of time in seconds spent in readdir operations.
gpfs_vfsx_readdir_tmax: Maximum amount of time in seconds spent in readdir operations.
gpfs_vfsx_readdir_tmin: Minimum amount of time in seconds spent in readdir operations.
gpfs_vfsx_readlink: Number of readlink operations.
gpfs_vfsx_readlink_t: Amount of time in seconds spent in readlink operations.
gpfs_vfsx_readlink_tmax: Maximum amount of time in seconds spent in readlink operations.
gpfs_vfsx_readlink_tmin: Minimum amount of time in seconds spent in readlink operations.
gpfs_vfsx_readpage: Number of readpage operations.
gpfs_vfsx_readpage_t: Amount of time in seconds spent in readpage operations.
gpfs_vfsx_readpage_tmax: Maximum amount of time in seconds spent in readpage operations.
gpfs_vfsx_readpage_tmin: Minimum amount of time in seconds spent in readpage operations.
gpfs_vfsx_remove: Number of remove operations.
gpfs_vfsx_remove_t: Amount of time in seconds spent in remove operations.
gpfs_vfsx_remove_tmax: Maximum amount of time in seconds spent in remove operations.
gpfs_vfsx_remove_tmin: Minimum amount of time in seconds spent in remove operations.
gpfs_vfsx_removexattr: Number of removexattr operations.
gpfs_vfsx_removexattr_t: Amount of time in seconds spent in removexattr operations.
gpfs_vfsx_removexattr_tmax: Maximum amount of time in seconds spent in removexattr operations.
gpfs_vfsx_removexattr_tmin: Minimum amount of time in seconds spent in removexattr operations.
gpfs_vfsx_rename: Number of rename operations.
gpfs_vfsx_rename_t: Amount of time in seconds spent in rename operations.
gpfs_vfsx_rename_tmax: Maximum amount of time in seconds spent in rename operations.
gpfs_vfsx_rename_tmin: Minimum amount of time in seconds spent in rename operations.
gpfs_vfsx_rmdir: Number of rmdir operations.
gpfs_vfsx_rmdir_t: Amount of time in seconds spent in rmdir operations.
gpfs_vfsx_rmdir_tmax: Maximum amount of time in seconds spent in rmdir operations.
gpfs_vfsx_rmdir_tmin: Minimum amount of time in seconds spent in rmdir operations.
gpfs_vfsx_setacl: Number of setacl operations.
gpfs_vfsx_setacl_t: Amount of time in seconds spent in setacl operations.
gpfs_vfsx_setacl_tmax: Maximum amount of time in seconds spent in setacl operations.
gpfs_vfsx_setacl_tmin: Minimum amount of time in seconds spent in setacl operations.
gpfs_vfsx_setattr: Number of setattr operations.
gpfs_vfsx_setattr_t: Amount of time in seconds spent in setattr operations.
gpfs_vfsx_setattr_tmax: Maximum amount of time in seconds spent in setattr operations.
gpfs_vfsx_setattr_tmin: Minimum amount of time in seconds spent in setattr operations.
gpfs_vfsx_setxattr: Number of setxattr operations.
gpfs_vfsx_setxattr_t: Amount of time in seconds spent in setxattr operations.
gpfs_vfsx_setxattr_tmax: Maximum amount of time in seconds spent in setxattr operations.
gpfs_vfsx_setxattr_tmin: Minimum amount of time in seconds spent in setxattr operations.
gpfs_vfsx_statfs: Number of statfs operations.
gpfs_vfsx_statfs_t: Amount of time in seconds spent in statfs operations.
gpfs_vfsx_statfs_tmax: Maximum amount of time in seconds spent in statfs operations.
gpfs_vfsx_statfs_tmin: Minimum amount of time in seconds spent in statfs operations.
gpfs_vfsx_symlink: Number of symlink operations.
gpfs_vfsx_symlink_t: Amount of time in seconds spent in symlink operations.
gpfs_vfsx_symlink_tmax: Maximum amount of time in seconds spent in symlink operations.
gpfs_vfsx_symlink_tmin: Minimum amount of time in seconds spent in symlink operations.
gpfs_vfsx_sync: Number of sync operations.
gpfs_vfsx_sync_t: Amount of time in seconds spent in sync operations.
gpfs_vfsx_sync_tmax: Maximum amount of time in seconds spent in sync operations.
gpfs_vfsx_sync_tmin: Minimum amount of time in seconds spent in sync operations.
gpfs_vfsx_tsfattr: Number of tsfattr operations.
gpfs_vfsx_tsfattr_t: Amount of time in seconds spent in tsfattr operations.
gpfs_vfsx_tsfattr_tmax: Maximum amount of time in seconds spent in tsfattr operations.
gpfs_vfsx_tsfattr_tmin: Minimum amount of time in seconds spent in tsfattr operations.
gpfs_vfsx_tsfsattr: Number of tsfsattr operations.
gpfs_vfsx_tsfsattr_t: Amount of time in seconds spent in tsfsattr operations.
gpfs_vfsx_tsfsattr_tmax: Maximum amount of time in seconds spent in tsfsattr operations.
gpfs_vfsx_tsfsattr_tmin: Minimum amount of time in seconds spent in tsfsattr operations.
gpfs_vfsx_unmap: Number of unmap operations.
gpfs_vfsx_unmap_t: Amount of time in seconds spent in unmap operations.
gpfs_vfsx_unmap_tmax: Maximum amount of time in seconds spent in unmap operations.
gpfs_vfsx_unmap_tmin: Minimum amount of time in seconds spent in unmap operations.
gpfs_vfsx_vget: Number of vget operations.
gpfs_vfsx_vget_t: Amount of time in seconds spent in vget operations.
gpfs_vfsx_vget_tmax: Maximum amount of time in seconds spent in vget operations.
gpfs_vfsx_vget_tmin: Minimum amount of time in seconds spent in vget operations.
gpfs_vfsx_write: Number of write operations.
gpfs_vfsx_write_t: Amount of time in seconds spent in write operations.
gpfs_vfsx_write_tmax: Maximum amount of time in seconds spent in write operations.
gpfs_vfsx_write_tmin: Minimum amount of time in seconds spent in write operations.
gpfs_vfsx_writepage: Number of writepage operations.
gpfs_vfsx_writepage_t: Amount of time in seconds spent in writepage operations.
gpfs_vfsx_writepage_tmax: Maximum amount of time in seconds spent in writepage operations.
gpfs_vfsx_writepage_tmin: Minimum amount of time in seconds spent in writepage operations.

Note: The GPFSVFS sensor cannot be used for IBM Storage Scale 5.1.0 and later versions. When the IBM Storage Scale is upgraded to IBM Storage Scale 5.1.0, the user needs to perform the following manual steps to migrate from the GPFSVFS sensor to the GPFSVFSX sensor:

The GPFSVFSX sensor must be added to the performance monitoring configuration by using the following command:
```
mmperfmon config add --sensors /opt/IBM/zimon/defaults/ZIMonSensors_GPFSVFSX.cfg
```
The GPFSVFS sensor must be disabled by using the following command:
```
mmperfmon config update GPFSVFS.period=0
```

External calls to the mmpmon vfssx interfere with the GPFSVFSX sensor. As the mmpmon command resets the vfssx metrics after every call, it might cause both the GPFSVFSX sensor and the external caller to retrieve inaccurate data.

Note:

The GPFSVFSX read and write operation time metrics do not provide meaningful minimum and maximum values. The gpfs_vfsx_read_tmin and gpfs_vfsx_read_tmax have the same values as the summarized time, which is gpfs_vfsx_read_t. Similarly, the gpfs_vfsx_write_tmin, gpfs_vfsx_write_tmax, and gpfs_vfsx_write_t have the same values.

This is due to the aggregation of various read sub-operations, such as gpfs_f_read, gpfs_f_readv, and gpfs_f_aio_read (that is completed synchronously) into gpfs_vfsx_read. And also, due to the aggregation of various write sub-operations, such as gpfs_f_write, gpfs_f_writev, and gpfs_f_aio_write (that is completed synchronously) into gpfs_vfsx_write.

GPFSVIO64

These metrics provide details of the virtual I/O server (VIOS) operations, where VIOS is supported.

Note: GPFSVIO64 is a replacement for GPFSVIO sensor and uses 64-bit counters.

gpfs_vio64_fixitOps: Number of VIO fix strips operations with read medium error.
gpfs_vio64_flushUpWrOps: Numbers of VIO flush update operations.
gpfs_vio64_flushPFTWOps: Numbers of VIO flush promoted full-track write operations.
gpfs_vio64_forceConsOps: Number of VIO force consistency operations.
gpfs_vio64_FTWOps: Numbers of VIO full-track write operations.
gpfs_vio64_logTipReadOps: Number of VIO log-tip read operations.
gpfs_vio64_logHomeReadOps: Number of VIO log-home read operations.
gpfs_vio64_logWriteOps: Number of VIO logs write operations.
gpfs_vio64_medWriteOps: Number of VIO mediums write operations.
gpfs_vio64_metaWriteOps: Number of recovery group metadata write operations.
gpfs_vio64_migrateTrimOps: Number of migrate trim operations.
gpfs_vio64_migratedOps: Number of VIO strip migration operations.
gpfs_vio64_promFTWOps: Number of VIO promoted full-track write operations.
gpfs_vio64_ptrackTrimOps: Number of ptrack trim operations.
gpfs_vio64_readCacheHit: Numbers of cache hit with read operations.
gpfs_vio64_readCacheMiss: Number of caches miss with read operations.
gpfs_vio64_readOps: Number of VIO read operations.
gpfs_vio64_RGDWriteOps: Number of recovery group descriptor write operations.
gpfs_vio64_scrubOps: Number of VIO scrub operations.
gpfs_vio64_shortWriteOps: Numbers of VIO short write operations.
gpfs_vio64_vtrackTrimOps: Number of vtrack trim operations.

Note: To report the new sensor data, the pmcollector must have the same or a higher code version than the pmsensors module. Otherwise, it ignores the data of the new sensors.

GPFSWaiters

For each independent fileset in the file system: Node- GPFSWaiters - waiters_time_threshold (all, 0.1s, 0.2s, 0.5s, 1.0s, 30.0s, 60.0s).

Note: Here 'all' implies a waiting time greater than or equal to 0 seconds.

For example, myNode|GPFSWaiters|all|gpfs_wt_count_all.

gpfs_wt_count_all: Count of all threads with waiting time greater than or equal to waiters_time_threshold seconds.
gpfs_wt_count_delay: Count of threads that are waiting for delay interval expiration with waiting time greater than or equal to waiters_time_threshold seconds.
gpfs_wt_count_local_io: Count of threads that are waiting for local I/O with waiting time greater than or equal to waiters_time_threshold seconds.
gpfs_wt_count_network_io: Count of threads that are waiting for network I/O with waiting time greater than or equal to waiters_time_threshold seconds.
gpfs_wt_count_syscall: Count of threads that are waiting for system call completion with waiting time greater than or equal to waiters_time_threshold seconds.
gpfs_wt_count_thcond: Count of threads that are waiting for a GPFS condition variable to be signaled with waiting time greater than or equal to waiters_time_threshold seconds.
gpfs_wt_count_thmutex: Count of threads that are waiting to lock a GPFS mutex with waiting time greater than or equal to waiters_time_threshold seconds.

GPFSFilesetQuota

The following metrics provide details of a fileset quota:

gpfs_rq_blk_current: Number of kilobytes currently in use.
gpfs_rq_blk_soft_limit: Assigned soft quota limit.
gpfs_rq_blk_hard_limit: Assigned hard quota limit.
gpfs_rq_blk_in_doubt: Number of kilobytes in-doubt, availability not yet resolved.
gpfs_rq_file_current: Number of files (inodes) currently in use.
gpfs_rq_file_soft_limit: Assigned soft quota limit.
gpfs_rq_file_hard_limit: Assigned hard quota limit.
gpfs_rq_file_in_doubt: Number of files (inodes) in-doubt, availability not yet resolved.

GPFSBufMgr

The following metric provides the current size of a page pool:

gpfs_bufm_tot_poolSizeK: Total size of the page pool.

Note: To activate the sensor on upgraded systems, run the mmperfmon config add --sensors /opt/IBM/zimon/defaults/ZIMonSensors_GPFSBufMgr.cfg command.

GPFSRPCS

Each metric that the GPFSRPCS sensor provides for a node is combined over all the peers to which the node is connected. An average is a weighted average over all the peer connections. A minimum is the minimum of the minimum for all the peer connections, and a maximum is the maximum of the maximum for all the peer connections.

The GPFSRPCS metrics are:

gpfs_rpcs_chn_av: The average amount of time the RPC must wait for access to a communication channel to the target node.
gpfs_rpcs_chn_cnt: The number of local nodes to this cluster (peers) this stat is collected from.
gpfs_rpcs_chn_max: The maximum amount of time the RPC must wait for access to a communication channel to the target node.
gpfs_rpcs_chn_min: The minimum amount of time the RPC must wait for access to a communication channel to the target node.
gpfs_rpcs_lat_tcp_av: The average latency of the RPC when sent and received over an Ethernet interface.
gpfs_rpcs_lat_tcp_cnt: The number of local nodes to this cluster (peers) this stat is collected from.
gpfs_rpcs_lat_tcp_max: The maximum latency of the RPC when sent and received over an Ethernet interface.
gpfs_rpcs_lat_tcp_min: The minimum latency of the RPC when sent and received over an Ethernet interface.
gpfs_rpcs_rcv_tcp_av : The average amount of time to transfer an RPC message from an Ethernet interface into the daemon.
gpfs_rpcs_rcv_tcp_cnt: The number of local nodes to this cluster (peers) this stat is collected from.
gpfs_rpcs_rcv_tcp_max: The maximum amount of time to transfer an RPC message from an Ethernet interface into the daemon.
gpfs_rpcs_rcv_tcp_min: The minimum amount of time to transfer an RPC message from an Ethernet interface into the daemon.
gpfs_rpcs_snd_tcp_av: The average amount of time to transfer an RPC message to an Ethernet interface.
gpfs_rpcs_snd_tcp_cnt: The number of local nodes to this cluster (peers) this stat is collected from.
gpfs_rpcs_snd_tcp_max: The maximum amount of time to transfer an RPC message to an Ethernet interface.
gpfs_rpcs_snd_tcp_min: The minimum amount of time to transfer an RPC message to an Ethernet interface.

Computed Metrics

These metrics can be used only through the mmperfmon query command. The following metrics are computed for GPFS:

gpfs_create_avg_lat (latency): gpfs_vfs_create_t / gpfs_vfs_create
gpfs_read_avg_lat (latency): gpfs_vfs_read_t / gpfs_vfs_read
gpfs_remove_avg_lat (latency): gpfs_vfs_remove_t / gpfs_vfs_remove
gpfs_write_avg_lat (latency): gpfs_vfs_write_t / gpfs_vfs_write

Important: The performance monitoring information that is driven by the IBM Storage Scale's internal monitoring tool and driven by the users who use the mmpmon command might affect each other.