• Monitor system performance using the CLI

    PDF

    Monitor system performance using the CLI

    Working with performance metrics in the Unisphere CLI

    You can use the Unisphere CLI to enable/disable the metrics collection service, view available metrics paths, and view historical and real-time metrics values.

    The metrics collection service is enabled by default in the CLI.

    To view metric values in the CLI, you must specify a metrics path. The following table summarizes the CLI commands used to display paths for the available metrics:

    Command
    Description
    / metrics/metric show
    Displays all available metrics paths.
    /metrics/metric -availability historical show
    Displays the paths for all available historical metrics.
    /metrics/metric -availability real-time show
    Displays the paths for all available real-time metrics.

    For information about using the CLI to enable or disable metrics settings and to display specific metrics, see the following sections.

    Manage metrics service

    Storage system metrics gather information about system performance and storage usage, and collect that information for user review. Analyzing the system metrics can help predict the future growth of the system.

    Historical and real-time metrics values are available in predefined intervals. High frequency (short interval) metric values are not kept as long as low frequency (long interval) metrics.

    The following table lists the metrics service attributes:

    Table 1. Metrics service attributes
    Attribute
    Description
    History enabled
    Indicates whether historical metrics collection is enabled. Value is one of the following:
    • yes
    • no
    Default value is yes.
    History retention
    Identifies the timestamp of the earliest available value for each frequency interval. The formats are:
    • YYYY-MM-DD HH:MM:SS (60 sec)
    • YYYY-MM-DD HH:MM:SS (300 sec)
    • YYYY-MM-DD HH:MM:SS (3600 sec)
    • YYYY-MM-DD HH:MM:SS (14400 sec)
    If the data for a certain interval is not available, the system displays not available instead of a timestamp.
    By default, the timestamps are UTC time. If you specify a timezone offset with -gmtoff, the timestamps adjust accordingly.

    View metrics service settings

    View the current metrics service settings.

    Use the show action command to change the output format.
    Format
    /metrics/service show
    Example

    The following command displays the metrics service settings for the system:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/service show 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection


1:  History enabled  = yes
    History retention= 2012-9-20 12:00:00 (60 sec), 2012-9-14 12:00:00 (300 sec), not available (3600 sec), not available (14400 sec)

    Configure metrics service

    Enable historical metrics collection.

    Format
    /metrics/service set -historyEnabled { yes | no }
    Only administrators are allowed to run this command.
    Action qualifiers
    Qualifier
    Description
    -historyEnabled
    Indicates whether historical metrics collection is enabled or disabled. Value is one of the following:
    • yes
    • no
    The system prompts for confirmation if you specify no.
    Example

    The following command enables metrics collection:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/service set -historyEnabled yes 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection


Operation completed successfully.

    Manage metrics settings

    Storage system metrics gather information about system performance and storage usage, and collect that information for user review. Analyzing the system metrics can help predict the future growth of the system.

    The following table lists the metrics attributes:

    Table 2. Metrics attributes
    Attribute
    Description
    Path
    Unique ID for the metric.
    Metrics are usually associated with objects. This association is reflected by a * character in the metric path, such as sp.*.net.device.*.bytes, which is associated with two objects, SP and network device. The metrics commands will accept a metric path with the * replaced by an object, and return only the result for the specified object. The system generates an error if the specified object is not valid.
    Description
    Description of the metric.
    Type
    Metric type. Valid values are:
    • rate — A counter difference relative to a unit of time.
    • counter — A monotonically increasing, unsigned quantity.
    • fact — Represents point-in-time information. Fact values should be expected to go up and down.
    • 64 bits counter — A counter of 64 bits.
    • text — Literal.
    Unit
    Unit measure for the metric.
    Availability
    Availability of the metric. Value is one of the following:
    • historical — The metric is included in historical metrics collection.
    • real-time — The metric supports real-time subscription.
    • historical, real-time — The metric supports both historical and real-time collection.
    This attribute does not apply to family, set, and compound metrics.

    View metrics settings

    View information about supported metrics.

    Use the show action command to change the output format.
    Format
    /metrics/metric [-path <value>] [-availability { historical | real-time } ] show
    Object qualifier
    Qualifier
    Description
    -path
    Specify a comma-separated list of metric paths.
    When typing metric paths, replace . with \., , with \, and \ with \\ in the object names.
    Omitting this switch specifies all available metrics.
    -availability
    Specify a type of metric to display. Value is one of the following:
    • historical
    • real-time
    Omitting this switch displays all metrics.
    Example 1

    The following command displays all available metric service settings for the system:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/metric show 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

1:  Path         = sp.*.cifs.global.basic.readsRate

2:  Path         = sp.*.cifs.global.basic.totalCallsRate

3:  Path         = sp.*.cifs.global.basic.writeAvgSize
    Example 2

    The following command displays all available metric service settings for the system with additional details:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/metric show -detail 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

1:    Path         = sp.*.blockCache.global.summary.cleanPages
      Description  = Number of Clean Pages on SP, based on a logical 
						               64 KB page size
      Type         = fact
      Unit         = Count
      Availability = real-time

2:    Path         = sp.*.blockCache.global.summary.dirtyBytes
      Description  = Amount of Dirty Data (MB) on SP
      Type         = fact
      Unit         = MB
      Availability = historical, real-time

3:    Path         = sp.*.blockCache.global.summary.dirtyPages
      Description  = Number of Dirty Pages on SP, based on a logical 
                     64 KB page size
      Type         = fact
      Unit         = Count
      Availability = real-time
    Example 3

    The following command displays all available real-time metric service settings for the system:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/metric -availability real-time show 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

439:  Path = sp.*.storage.pool.*.sizeTotal

440:  Path = sp.*.storage.pool.*.sizeUsed

441:  Path = sp.*.storage.pool.*.sizeUsedBlocks

442:  Path = sp.*.storage.pool.*.snapshotSizeSubscribed

443:  Path = sp.*.storage.pool.*.snapshotSizeUsed

444:  Path = sp.*.storage.summary.readBlocksRate

445:  Path = sp.*.storage.summary.readBytesRate

446:  Path = sp.*.storage.summary.readsRate

447:  Path = sp.*.storage.summary.totalBytesRate
    Example 4

    The following command displays the metrics service settings for the metrics with the specified paths:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/metric -path sp.*.storage.lun.*.avgReadSize,sp.*.storage.filesystem.*.writesRate,sp.*.cifs.smb2.basic.readsRate show -detail 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

1:  Path         = sp.*.storage.lun.*.avgReadSize
    Description  = Average read size on this LUN
    Type         = fact
    Unit         = KB
    Availability = historical, real-time

2:  Path         = sp.*.storage.filesystem.*.writesRate
    Description  = Rate of sp.*.storage.filesystem.*.writes
    Type         = rate
    Unit         = Requests/s
    Availability = historical, real-time

3:  Path         = sp.*.cifs.smb2.basic.readsRate
    Description  = Rate of sp.*.cifs.smb2.basic.reads
    Type         = rate
    Unit         = Ops/s
    Availability = real-time

    Manage historical metrics values

    Storage system metrics gather information about system performance and storage usage, and collect that information for user review. Analyzing the system metrics can help predict the future growth of the system.

    Historical metric values are available in predefined intervals. High frequency (short interval) metric values are not kept as long as low frequency (long interval) metrics.

    The following table lists the historical metrics attributes:

    Table 3. Historical metrics attributes
    Attribute
    Description
    Timestamp
    Time when the metric value was collected. The format is: YYYY-MM-DD HH:MM:SS, where:
    • YYYY — Year
    • MM — Month
    • DD — Day
    • HH — Hour
    • MM — Minute
    • SS — Second
    Dynamic attributes
    Identifies the object name or metric value.

    View historical metrics settings

    View historical metrics settings. The default output appears in a tabular format.

    Use the show action command to change the output format.
    Format
    /metrics/value/hist -path <value> show -interval { 60 | 300 | 3600 | 14400 }[ -from <value> ] [ -to <value>] [ -count <value> ][ -flat ][ -summary ]
    Object qualifier
    Qualifier
    Description
    -path
    Specify a comma-separated list of metric paths.
    When typing metric paths, replace . with \., , with \, and \ with \\ in the object names.
    -interval
    Specify an interval for the metric values. Default interval is seconds.
    -from
    Specify the start of the query period. The format is: YYYY-MM-DD HH:MM:SS or YYYY-MM-DDTHH:MM:SS, where:
    • YYYY — Year
    • MM — Month
    • DD — Day
    • T — Time delimiter
    • HH — Hour
    • MM — Minute
    • SS — Second
    Ensure that the value is a time in the past. You can choose to specify just the date (in the YYYY-MM-DD format) or the time (in the HH:MM:SS format). If you do not specify the time, the system automatically uses 00:00:00. If you choose to not specify the date, the current system date is used.
    -to
    Specify the end of the query period. The format is: YYYY-MM-DD HH:MM:SS or YYYY-MM-DDTHH:MM:SS, where:
    • YYYY — Year
    • MM — Month
    • DD — Day
    • T — Time delimiter
    • HH — Hour
    • MM — Minute
    • SS — Second
    Ensure that the value is a time in the past. You can choose to specify just the date (in the YYYY-MM-DD format) or the time (in the HH:MM:SS format). If you do not specify the time, the system automatically uses 00:00:00. If you choose to not specify the date, the current system date is used.
    -count
    Specify the number of samples to display. A sample is a set of metric values related to a single timestamp. Valid values are numbers greater than or equal to one.
    -flat
    Displays the member values for grouped metrics.
    -summary
    Displays the maximum, minimum, and average value for each metric.
    The -from and -to qualifiers take precedence over the -count qualifier. In the example below, only 7 samples exist between the from and to dates. Although the value for the -count qualifier is set to 10, only 7 values appear. If the -from and -to qualifiers are not specified, the output will include 10 samples.
    Examples of output with different combinations of the -from, -to, and -count qualifiers

    The following table illustrates the output that appears with combinations of the -from, -to, and -count qualifiers. It assumes that the current time is 2012-09-21 12:30:00.

    Qualifier Combination
    Output
    -from <future date/time>
    Example: -from “2012-09-21 12:31:00”

    Result: This results in an error because the time for the -from qualifier is specified in the future.

    -from <current date/time or date/time in the past>

    -to <future date/time>

    Example: -from “2012-09-01 00:00:00” -to “2012-09-21 12:31:00”

    Result: This results in an error because the time for the -to qualifier is specified in the future.

    -from <date/time in the past> -count <value>
    Example: -from “2012-09-20 01:02:00” -count 100

    Result: The result includes 100 samples from “2012-09-20 01:02:00”. If there are less than 100 samples available, the result lists all samples from the specified time to the current time.

    -from <date/time in the past>

    -to <current date/time or date/time in the past>

    -count <value>

    Example: -from “2012-09-20 01:02:00” -to “20-09-20 12:00:00” -count 100

    Result: The result includes 100 samples within the specified time period. If there are less than 100 samples available, the result lists all samples within the time period.

    -to <current date/time or date/time in the past>

    -count <value>

    Example: -to “20-09-20 12:00:00” -count 100

    Result: The result includes the latest 100 samples before the specified time. If there are less than 100 samples available, the result lists all samples.

    -count <value>
    Example: -count 100

    Result: The result includes the latest 100 samples, or if there are less than 100 samples available, the result lists all samples.

    -to <current date/time or date/time in the past>
    Example: -to “20-09-20 12:00:00”

    Result: The result includes all samples from the timestamp of the earliest sample to the specified time.

    -from, -to, and -count are not specified.
    Result: The result includes the latest 100 samples, or if there are less than 100 samples available, the result lists all samples. This is equivalent to “ -count 100”.
    Example 1

    The following command displays the specified individual metric SPA LUN sv_1 every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.spa.storage.lun.sv_1.readsRate show -interval 60 -from "2014-06-24 02:12:00" -to "2014-06-24 02:1 4 :00" 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP        |LUN               |Read
                    |          |                  |Counts/s
--------------------+----------+------------------+--------
2014-06-24 02:12:00 |spa       |sv_1              |   4.001
2014-06-24 02:13:00 |spa       |sv_1              |   2.400
2014-06-24 02:14:00 |spa       |sv_1              |   9.602
    Example 2

    The following command displays the specified metric, associated with a single object type, SPs, every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.*.cpu.summary.utilization show -interval 60 -from "2014-06-24 02:57:00" -to "2014-06-24 02:59:10" 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP        |summary
                    |          |CPU
                    |          |Util %
--------------------+----------+-------
2014-06-24 02:57:00 |spa       |  12.62
                    |spb       |  32.46
2014-06-24 02:58:00 |spa       |  13.06
                    |spb       |  19.75
2014-06-24 02:59:00 |spa       |  13.44
                    |spb       |  32.47
    Example 3

    The following command displays the specified metric, associated with two object types, SPs and LUNs, every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.*.storage.lun.*.readsRate show -interval 60 -from "2014-06-24 02:59:00" -to "2014-06-24 03:01:00" 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP        |LUN               |Read
                    |          |                  |Counts/s
--------------------+----------+------------------+--------
2014-06-24 02:59:00 |spa       |sv_1              |   0.050
                    |spa       |sv_2              |       0
                    |spb       |sv_1              |       0
                    |spb       |sv_2              |   0.033
2014-06-24 03:00:00 |spa       |sv_1              |   0.467
                    |spa       |sv_2              |       0
                    |spb       |sv_1              |       0
                    |spb       |sv_2              |   0.117
2014-06-24 03:01:00 |spa       |sv_1              |   0.833
                    |spa       |sv_2              |       0
                    |spb       |sv_1              |       0
                    |spb       |sv_2              |   0.467
    Example 4

    The following command displays the specified metric, associated with three object types, SPs, pools, and LUNs, every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.*.storage.pool.*.lun.*.dataSizeAllocated show -interval 60 -from "2014-06-24 03:04:00" -to "2014-06-24 03:06:00" 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP  |Pool Statistics |LUN   |Data Size Allocated Bytes
--------------------+----+----------------+------+--------------------------
2014-06-24 03:04:00 |spa |pool_1          |sv_1  |6442450944
                    |spa |pool_1          |sv_2  |8589934592
                    |spb |pool_1          |sv_1  |6442450944
                    |spb |pool_1          |sv_2  |8589934592
2014-06-24 03:05:00 |spa |pool_1          |sv_1  |6442450944
                    |spa |pool_1          |sv_2  |8589934592
                    |spb |pool_1          |sv_1  |6442450944
                    |spb |pool_1          |sv_2  |8589934592
2014-06-24 03:06:00 |spa |pool_1          |sv_1  |6442450944
                    |spa |pool_1          |sv_2  |8589934592
                    |spb |pool_1          |sv_1  |6442450944
                    |spb |pool_1          |sv_2  |8589934592
    Example 5

    The following command displays metrics, associated with two object types, SPs and LUNs, and an individual metric associated with SPA, every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.*.storage.lun.*.readsRate, sp.*.storage.lun.*.writesRate, sp. spa.cpu.summary.utilization show -interval 60 -from "2014-06-24 03:04:00" -to "2014-06-24 03:06:00" 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP     |LUN     |Read     |Write    |SP    |summary
                    |       |        |Counts/s |Counts/s |      |CPU
                    |       |        |          |        |      |Util %
--------------------+-------+--------+------+--------+----------+-------
2014-06-24 03:10:00 |spa    |sv_1    |        0|        0|spa   |  12.63
                    |spa    |sv_2    |    1.050|    9.066|      | 
                    |spb    |sv_1    |    0.067|    9.350|      |
                    |spb    |sv_2    |    0.100|    14.95|      |
2014-06-24 03:11:00 |spa    |sv_1    |        0|        0|spa   |  12.56
                    |spa    |sv_2    |    0.700|    26.62|      |
                    |spb    |sv_1    |    0.167|    12.28|      |
                    |spb    |sv_2    |    2.883|    25.65|      |
2014-06-24 03:12:00 |spa    |sv_1    |    0.667|    19.53|spa   |  12.12
                    |spa    |sv_2    |    0.333|    26.87|      |
                    |spb    |sv_1    |    7.066|    3.700|      |
                    |spb    |sv_2    |    7.066|    3.383|      |
    Example 6

    The following command displays the member values for specified metrics every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.*.cpu.summary.utilization show -interval 60 -from "2014-06-24 03:14:00" -to "2014-06-24 03:16:00" -flat 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |spa    |spb
                    |summary|summary
                    |CPU    |CPU
                    |Util % |Util %
--------------------+-------+-------
2014-06-24 03:14:00 |  15.06|  26.78
2014-06-24 03:15:00 |  15.82|  29.39
2014-06-24 03:16:00 |  15.94|  23.59
    Example 7

    The following command displays the maximum, minimum, and average value for each metric every 60 seconds during the query period:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! /metrics/value/hist -path sp.*.cpu.summary.utilization show -interval 60 -from "2014-06-24 03:19:00" -to "2014-06-24 03:21:00" -summary 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP        |summary
                    |          |CPU
                    |          |Util %
--------------------+----------+-------
2014-06-24 03:19:00 |spa       |  17.72
                    |spb       |  43.52
2014-06-24 03:20:00 |spa       |  15.35
                    |spb       |  37.82
2014-06-24 03:21:00 |spa       |  15.08
                    |spb       |  36.32

Summary             |SP        |summary
                    |          |CPU
                    |          |Util %
--------------------+----------+-------
Minimum             |spa       |  15.08
                    |spb       |  36.32
Average             |spa       |  16.05
                    |spb       |  39.22
Maximum             |spa       |  17.72
                    |spb       |  43.52

    Manage real-time metrics values

    Storage system metrics gather information about system performance and storage usage, and collect that information for user review. Analyzing the system metrics can help predict the future growth of the system.

    The following table lists the real-time metrics attributes.

    Table 4. Real-time metrics attributes
    Attribute
    Description
    Timestamp
    Time when the metric value was collected. The format is: YYYY-MM-DD HH:MM:SS, where:
    • YYYY — Year
    • MM — Month
    • DD — Day
    • HH — Hour
    • MM — Minute
    • SS — Second
    Dynamic attributes
    Identifies the object name or metric value.

    View real-time metrics settings

    View real-time metrics settings. The default output appears in a tabular format.

    Use the show action command to change the output format.
    Format
    /metrics/value/rt -path <value> show -interval <value> [ -to <value>] [ -count <value> ][ -flat ][ -summary ]
    Object qualifier
    Qualifier
    Description
    -path
    Specify a comma-separated list of metric paths.
    When typing metric paths, replace . with \., , with \, and \ with \\ in the object names.
    Action qualifier
    Qualifier
    Description
    -interval
    Specify an interval for the metric values. Default interval is seconds.
    -to
    Specify the end of the query period. The format is: YYYY-MM-DD HH:MM:SS or YYYY-MM-DDTHH:MM:SS, where:
    • YYYY — Year
    • MM — Month
    • DD — Day
    • T — Time delimiter
    • HH — Hour
    • MM — Minute
    • SS — Second
    Ensure that the value is a time in the past. You can choose to specify just the date (in the YYYY-MM-DD format) or the time (in the HH:MM:SS format). If you do not specify the time, the system automatically uses 00:00:00. If you choose to not specify the date, the current system date is used.
    -count
    Specify the number of samples to display. A sample is a set of metric values related to a single timestamp. Valid values are numbers greater than or equal to one.
    -flat
    Displays the member values for grouped metrics.
    -summary
    Displays the maximum, minimum, and average value for each metric.
    Objects can come and go at any time, mostly due to object creation and deletion. In flat format, every time a new object is included, the title in tabular or CSV format or the attributes in NVP format is adjusted accordingly and reprinted as necessary on screen. If an object is no longer valid but it already has a column in tabular or CSV format, the column is kept only if its value becomes blank. Otherwise the object is not displayed anymore.
    Example 1

    The following command displays the specified real-time metric every 10 seconds:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! uemcli /metrics/value/rt -path sp.*.storage.lun.*.readsRate show -interval 10 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp           |SP        |LUN               |Read
                    |          |                  |Counts/s
--------------------+----------+------------------+--------
2014-06-24 03:26:10 |spb       |sv_1              |   0.225
2014-06-24 03:26:20 |spb       |sv_1              |   0.200
                    |spb       |sv_2              |   0.100
2014-06-24 03:26:30 |spb       |sv_2              |   0.200
    Example 2

    The following command displays the member values for the specified grouped real-time metric every 10 seconds in comma-separated values (CSV) format:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! uemcli /metrics/value/rt -path sp.*.storage.lun.*.readsRate show -interval 10 -flat -output csv 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

Timestamp,spb sv_1 Read Counts/s 
2014-06-24 03:26:10,0.225
Timestamp,spb sv_1 Read Counts/s,spb sv_2 Read Counts/s
2014-06-24 03:26:20,0.200,0.100
2014-06-24 03:26:30,,0.200
    Example 3

    The following command displays the specified real-time metric every 10 seconds name-value pair (NVP) format:

    uemcli -d 10.0.0.1 -u Local/joe -p MyPassword456! uemcli /metrics/value/rt -path sp.*.storage.lun.*.readsRate show -interval 10 -output nvp 
                              Storage system address: 10.0.0.1
Storage system port: 443
HTTPS connection

1: Timestamp = 2014-06-24 03:26:10
   SP        = spb
   Client    = sv_1
   CIFS Read = 0.225

2: Timestamp = 2014-06-24 03:26:20
   SP        = spb
   Client    = sv_1
   CIFS Read = 0.200

3: Timestamp = 2014-06-24 03:26:20
   SP        = spb
   Client    = sv_2
   CIFS Read = 0.100

4: Timestamp = 2014-06-24 03:26:30
   SP        = spb
   Client    = sv_2
   CIFS Read = 0.200