Dell PowerVault ME5 Series Storage System CLI Reference Guide

Shows one or more collected data points for a list of metrics.

Before you can view metrics, you must start metric retention by using the start metrics command.

[calculate average|max|min|count|sum]

[count <number-of-data-samples>|all]

[database dynamic|historical]

[filename <filename>.csv]

[time-range "<date/time-range>"]

metrics-list

calculate average|max|min|count|sum

Optional. Instead of listing columns of metric data points, apply a calculation to each column and display a single, summary value for each column.

[count <number-of-data-samples>|all]

Optional. Specifies the number of data samples to display, from 1 to 5000, or all available samples. Each sample is shown as a separate row in the command output. If this parameter is omitted, 100 samples are shown. If you specify this parameter, do not specify the time-range parameter.

Due to memory consumption limits, it is recommended to use the all parameter with the filename parameter to direct output to a CSV file instead of to the console. If the filename parameter is not specified, the all parameter might display an error message requiring you to reduce the sample count or the number of requested metrics or both.

[database dynamic|historical]

Optional. Specifies whether to run the query on current dynamic data or historical data. The default is dynamic.

[filename <filename>.csv]

Optional. Saves metrics to a CSV file that is accessible from either controller module.

To access the file, use SFTP or FTP. Files are created in the metrics/ folder. Use the cd command to change to the metrics/ folder and the ls command to list folder contents. The folder is limited to 20 files and/or 100MB of total space consumed. As the folder fills, the system automatically deletes the oldest files to make space for new files.

[time-range "<date/time-range>"]

Optional. Specifies the date/time range of historical metrics to query, in the format "start yyyy-mm-dd hh:mm [AM|PM] end yyyy-mm-dd hh:mm [AM|PM]". If the start date/time is specified but no end date/time is specified, the current date/time is used as the end date/time. The system returns the oldest sample taken after the start time and the latest sample taken before the end time. If the specified start date/time is earlier than the oldest sample, that sample is used as the start date/time. If you specify this parameter, do not specify the count parameter. If both the count and time-range parameters are omitted, the most recent 100 data samples are displayed.

Times are saved at a strict cadence of :00, :05, :10 seconds, and so on. To display a single value at a particular time, ensure that the start and end times are identical and match a sample time exactly.

metrics-list

Specifies a comma-separated list of metrics whose data points you want to display. For each metric use the format typefield.instance, where: type is a type of storage object; field is a specific measured property of that object; and instance is the name or serial number of that object. If the instance value contains a space, the entire metric must be enclosed in double quotes. For example: "volume.read-iops.My Vol1". Do not include an instance value for system metrics.

If database historical is specified, you can optionally specify to show maximum, minimum, or average values. Each historical data point is calculated from a range of dynamic data points. The default data point calculation is based on the average of the dynamically sampled data points. However, historical data also retains minimum and maximum values for each calculated historical data point. To show minimum values append @min; to show maximum values append @max; to explictly show average values append @average. For example: "volume.read-iops@max.My Vol1".

• total-avg-response-time: Average response time of an operation in microseconds. Operations include both reads and writes. Applicable objects: controller, host-port, pool, system, volume.
• total-bytes-per-second: Sum of read bytes per second and write bytes per second. Applicable objects:controller, host-port, pool, system, volume.
• total-iops: Sum of read IOPS and write IOPS. Applicable storage objects:controller, host-port, pool, system, volume.
• total-max-response-time: Sum of read maximum response time and write maximum response time. Applicable storage objects:controller, host-port, pool, system, volume.
• total-num-bytes: Sum of read bytes and write bytes. Applicable objects: controller, host-port, pool, system, volume.
• read-io-count: Number of read I/O operations. Applicable objects: controller, host-port, pool, system, volume.
• read-ahead-ops: Number of times that read ahead pre-fetched data for host reads. Applicable objects: controller, volume.
• read-avg-queue-depth: Average number of pending read operations being serviced since the last sampling time. This value represents periods of activity only and excludes periods of inactivity. Applicable objects: host-port, volume.
• read-avg-response-time: I/O read average response time in microseconds. Applicable objects: controller, host-port, pool, system, volume.
• read-bytes-per-second: Number of bytes read per second. Applicable storage objects: controller, host-port, pool, system, volume.
• read-iops: Number of I/Os per second. Applicable objects: controller, host-port, pool, system, volume.
• read-max-response-time: Maximum I/O read response time in microseconds. Applicable objects: controller, host-port, pool, system, volume.
• read-num-bytes: Number of bytes read since the last time this data point was sampled. Applicable objects: controller, host-port, pool, system, volume.
• small-destages: Number of partial stripe destages. (These tend to be very inefficient compared to full stripe writes.) Applicable objects: controller,volume.
• write-io-count: Number of write I/O operations. Applicable objects: controller, host-port, pool, system, volume.
• rite-avg-queue-depth: Average number of pending write operations being serviced since the last sampling time. This value represents periods of activity only and excludes periods of inactivity. Applicable objects: host-port, volume.
• write-avg-response-time: I/O write average response time in microseconds. Applicable objects: controller, host-port, pool, system, volume.
• write-bytes-per-second: Number of bytes written per second. Applicable objects: controller, host-port, pool, system, volume.
• write-cache-space: Current size of write cache in 16KB chunks. Applicable objects: controller, volume.
• write-cache-percent: Percentage of write cache currently being used in tenths of a percent. Applicable objects: controller, volume.
• write-full-stripe-destages: Number of full stripe destages, which are the most efficient destage type. Applicable objects: controller, volume.
• write-iops: Number of I/Os per second. Applicable objects: controller, host-port, pool, system, volume.
• write-max-response-time: Maximum I/O write response time in microseconds. Applicable objects: controller, host-port, pool, system, volume.
• write-num-bytes: Number of bytes written since the last time this data point was sampled. Applicable objects: controller, host-port, pool, system, volume.

Metrics for the system storage object are synthesized from data captured by controller storage objects.

Column Key

For each specified metric, the numeric ID of the corresponding column in the displayed data table. Column IDs increment from 01.

Time

The date/time of data points in the same row.

01

The data points for the first specified metric.

<IDs>

The data points for other specified metrics, if any.

# query metrics count 10 system.read-avg-response-time

show metrics-list

start metrics

stop metrics