mmptop command

Reports top entities of IBM Storage Scale performance statistics (dynamic real-time view).

Synopsis

mmptop Metric[,Metric...] | NamedQuery | Measurement | +  [-N Nodename | all] [--bucket-size size] [--filter filter]

Availability

Available on all IBM Storage Scale editions. Available and needed only on Linux®.

Description

This command visualizes metrics that have multiple entities, such as pdisks, nodes, file system, and shows the top entities by printing one line per entity that is sorted by first and second metric. Any metric that is known to the pmcollector can be queried. The command runs a query and generates a report every interval (default 1 second).

Parameters

Metric[,Metric...]

Specifies a comma-separated list of metrics to display in the output.

NamedQuery

Specifies the name of a predefined query.

Measurement

Specifies the name of a predefined measurement and prints a list of known metrics.

* -N or --Node Nodename

Specifies the node from where the metrics are retrieved. Accepts a single node only or "all" for all nodes. For general information on how to specify node names, see Specifying nodes as input to GPFS commands.

+

Prints a list of known metrics.

--bucket-size size

Specifies the bucket size (number of seconds) for the query and the report interval, the default bucket size is 1 s.

If the bucket size is 10, then data is queried and reported every 10 s. If the bucket size is smaller than the sensor period of a metric, the reports are generated frequently but the metric remains empty.
If bucket size is 1 s but sensor period is 10 s, report shows every 10 s the metric value and all other reports remain empty.

--filter filter

Specifies the filter criteria for the query to run. Use this parameter to filter any values in the metric key definition.

Each metric has different key components. For example, the metric gpfs_nsdds_bytes_read is disk related and, therefore, filters by gpfs_disk_name. The metric gpfs_fs_bytes_written is per file system and, therefore, filters by gpfs_fs_name.

You can specify multiple filters by using a , (comma) to separate each filter.

To see a list of all the available filters, issue --filter +.

Exit status

0: Successful completion
1: Invalid options

Security

You must have root authority to run the mmptop command.

Examples

Query nodes with the highest CPU usage.
```
# mmptop cpu -N all
```
Query top pdisks with the highest read/write values on an IBM Storage Scale Systemor IBM Storage Scale Erasure Code Edition I/O node.
```
# mmptop gpfs_pdds_bytes_read,gpfs_pdds_bytes_written -N ess5kio2 -b 10
```

Query top file systems with the highest read/write IOPS, refresh every 10 seconds.

# mmptop gpfs_fs_bytes_written,gpfs_fs_bytes_read -b 10

A sample output is as follows:

mmptop gpfs_fs_bytes_written,gpfs_fs_bytes_read -b 10
2024-09-23 18:01:35.783361
------------------------------------------------------------------------------------------------
Key                           | gpfs_fs_bytes_written| gpfs_fs_bytes_read|
------------------------------------------------------------------------------------------------
ems5k|ScaleCluster1|ess5k16m  3000232|                102020|
ems5k|ScaleCluster1|ess5k1m   2021256|                30056|
ems5k|ScaleCluster1|ess5k4m   4096|                   0|
ems5k|ScaleCluster1|ess5k8m   0|                      0|

Location

/usr/lpp/mmfs/bin