NVIDIA MLNX-GW User Manual for NVIDIA Skyway Appliance v8.1.3000
NVIDIA MLNX-GW User Manual for NVIDIA Skyway Appliance v8.2.2200

MLNX-GW—Statistics and Alarms

stats alarm clear

stats alarm <alarm ID> clear

Clears alarm state.

Syntax Description

alarm ID

Alarms supported by the system, for example:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

Default

N/A

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats alarm cpu_util_indiv clear

Related Commands

show stats alarm

Notes

stats alarm enable

stats alarm <alarm-id> enable
no stats alarm <alarm-id> enable

Enables the alarm.
The no form of the command disables the alarm, notifications will not be received.

Syntax Description

alarm ID

Alarms supported by the system, for example:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

Default

The default is different per alarm-id

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats alarm cpu_util_indiv enable

Related Commands

show stats alarm

Notes

stats alarm event-repeat

stats alarm <alarm ID> event-repeat {single | while-not-cleared}
no stats alarm <alarm ID> event-repeat

Configures repetition of events from this alarm.
The no form of this command resets this parameter to its default.

Syntax Description

alarm ID

Alarms supported by the system, for example:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

single

Does not repeat events: only sends one event whenever the alarm changes state.

while-not-cleared

Repeats error events until the alarm clears.

Default

single

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats alarm cpu_util_indiv event-repeat single

Related Commands

show stats alarm

Notes

stats alarm {rising | falling}

stats alarm <alarm ID> {rising | falling} {clear-threshold | error-threshold} <threshold-value>

Configure alarms thresholds.

Syntax Description

alarm ID

Alarms supported by the system, for example:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

falling

Configures alarm for when the statistic falls too low.

rising

Configures alarm for when the statistic rises too high.

error-threshold

Sets threshold to trigger falling or rising alarm.

clear-threshold

Sets threshold to clear falling or rising alarm.

threshold-value

The desired threshold value, different per alarm.

Default

Default is different per alarm-id

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats alarm cpu_util_indiv falling clear-threshold 10

Related Commands

show stats alarm

Notes

Not all alarms support all four thresholds.

stats alarm rate-limit

stats alarm <alarm ID> rate-limit {count <count-type> <count> | reset | window <window-type> <duration>}

Configures alarms rate limit.

Syntax Description

alarm ID

Alarms supported by the system, for example:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

count-type

Long medium, or short count (number of alarms).

reset

Set the count and window durations to default values for this alarm.

window-type

Long medium, or short count, in seconds.

Default

Short window: 5 alarms in 1 hour
Medium window: 20 alarms in 1 day
Long window: 50 alarms in 7 days

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats alarm paging rate-limit window long 2000

Related Commands

show stats alarm

Notes

stats chd clear

stats chd <CHD ID> clear

Clears CHD counters.

Syntax Description

CHD ID

CHD supported by the system, for example:

  • cpu_util—CPU utilization: percentage of time spent

  • cpu_util_ave—CPU utilization average: percentage of time spent

  • cpu_util_day—CPU utilization average: percentage of time spent

  • disk_device_io_hour—storage device I/O read/write statistics for the last hour: bytes

  • disk_io—operating system aggregate disk I/O average (KB/sec)

  • fs_mnt_day—filesystem system usage average: bytes

  • fs_mnt_month—filesystem system usage average: bytes

  • fs_mnt_week—filesystem system usage average: bytes

  • intf_day—network interface statistics aggregation: bytes

  • intf_hour—network interface statistics (same as “interface” sample)

  • intf_util—aggregate network utilization across all interfaces

  • memory_day—average physical memory usage: bytes

  • memory_pct—average physical memory usage

  • paging—paging activity: page faults

  • paging_day—paging activity: page faults

Default

N/A

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats chd memory_day clear

Related Commands

show stats chd

Notes

stats chd enable

stats chd <chd-id> enable
no stats chd <chd-id> enable

Enables the CHD.
The no form of the command disables the CHD.

Syntax Description

chd-id

CHD supported by the system, for example:

  • cpu_util—CPU utilization: percentage of time spent

  • cpu_util_ave—CPU utilization average: percentage of time spent

  • cpu_util_day—CPU utilization average: percentage of time spent

  • disk_device_io_hour—storage device I/O read/write statistics for the last hour: bytes

  • disk_io—operating system aggregate disk I/O average: KB/sec

  • fs_mnt_day—filesystem system usage average: bytes

  • fs_mnt_month—filesystem system usage average: bytes

  • fs_mnt_week—filesystem system usage average: bytes

  • intf_day—network interface statistics aggregation: bytes

  • intf_hour— network interface statistics (same as “interface” sample)

  • intf_util— aggregate network utilization across all interfaces

  • memory_day—average physical memory usage: bytes

  • memory_pct—average physical memory usage

  • paging—paging activity: page faults

  • paging_day—paging activity: page faults

Default

Enabled

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats chd memory_day enable

Related Commands

show stats chd

Notes

stats chd compute time

stats chd <CHD ID> compute time {interval | range} <number of seconds>

Sets parameters for when this CHD is computed.

Syntax Description

CHD ID

Possible IDs:

  • cpu_util—CPU utilization: percentage of time spent

  • cpu_util_ave—CPU utilization average: percentage of time spent

  • cpu_util_day—CPU utilization average: percentage of time spent

  • disk_device_io_hour—storage device I/O read/write statistics for the last hour: bytes

  • disk_io—operating system aggregate disk I/O average: KB/sec

  • fs_mnt_day—filesystem system usage average: bytes

  • fs_mnt_month—filesystem system usage average: bytes

  • fs_mnt_week—filesystem system usage average: bytes

  • intf_day—network interface statistics aggregation: bytes

  • intf_hour—network interface statistics (same as “interface” sample)

  • intf_util—aggregate network utilization across all interfaces

  • memory_day—average physical memory usage: bytes

  • memory_pct—average physical memory usage

  • paging—paging activity: page faults

  • paging_day—paging activity: page faults

interval

Specifies calculation interval (how often to do a new calculation) in number of seconds

range

Specifies calculation range, in number of seconds

number of seconds

Number of seconds

Default

Different per CHD

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats chd memory_day compute time interval 120

Related Commands

show stats chd

Notes

stats export

stats export <format> <sample-id>

Exports collected information to a file. Can export extended "interface-ethernet", "interface-port-channel", "interface-mlag-port-channel"& "power" samples.

Syntax Description

sample-id

Sample name for which report file should be generated.

  • cpu_util—CPU utilization: milliseconds of time spent

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

format

Format of report file

Default

N/A

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats export csv memory

Related Commands

show stats sample

Notes

stats sample clear

stats sample <sample ID> clear

Clears sample history.

Syntax Description

sample ID

Possible sample IDs are:

  • congested

  • cpu_util—CPU utilization: milliseconds of time spent

  • disk_device_io—storage device I/O statistics

  • disk_io—operating system aggregate disk I/O: KB/sec

  • fan - Fan speed

  • fs_mnt_bytes—filesystem usage: bytes

  • fs_mnt_inodes—filesystem usage: inodes

  • interface—network interface statistics

  • intf_util—network interface utilization: bytes

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

  • power—power supply usage

  • power-consumption

  • temperature—modules temperature

Default

N/A

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats sample temperature clear

Related Commands

show stats sample

Notes

stats sample enable

stats sample <sample-id> enable
no states sample <sample-id> enable

Enables the sample.
The no form of the command disables the sample.

Syntax Description

sample-id

Possible sample IDs are:

  • congested

  • cpu_util—CPU utilization: milliseconds of time spent

  • disk_device_io—storage device I/O statistics

  • disk_io—operating system aggregate disk I/O: KB/sec

  • fan—fan speed

  • fs_mnt_bytes—filesystem usage: bytes

  • fs_mnt_inodes—filesystem usage: inodes

  • interface—network interface statistics

  • intf_util—network interface utilization: bytes

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

  • power—power supply usage

  • power-consumption

  • temperature—modules temperature

  • interface-ethernet—Ethernet counters statistics: counter units

  • interface-mlag-port-channel—MLAG counters statistics: counter units

  • interface-port-channel—LAG counters statistics: counter units

Default

Enabled

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats sample temperature enable

Related Commands

show stats sample

Notes

stats sample interval

stats sample <sample-id> interval [<interval>]
no stats sample <sample-id> interval [<interval>]

Sets the sampling interval between taking of sample records.
The no form of the command sets interval to default value.

Syntax Description

sample-id

Sample name for which report file should be generated.

  • congested

  • cpu_util—CPU utilization: milliseconds of time spent

  • disk_device_io—storage device I/O statistics

  • disk_io—operating system aggregate disk I/O: KB/sec

  • fan—fan speed

  • fs_mnt_bytes—filesystem usage: bytes

  • fs_mnt_inodes—filesystem usage: inodes

  • interface—network interface statistics

  • intf_util—network interface utilization: bytes

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

  • power—power supply usage

  • power-consumption

  • temperature—modules temperature

  • interface-ethernet—Ethernet counters statistics: counter units

  • interface-mlag-port-channel—MLAG counters statistics: counter units

  • interface-port-channel—LAG counters statistics: counter units

interval

Measured in seconds. Range: 1 - 86400 (24 hours)

Default

Default for “interface” samples is 60 seconds

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats sample interface-ethernet interval 1

Related Commands

show stats sample

Notes

stats sample max-entries

stats sample <sample-id> max-entries [<max-entries>]
no stats sample <sample-id> max-entries [<max-entries>]

Sets number of records to be kept in memory for the counter.
The no form of the command resets the value to its default.

Syntax Description

sample-id

Sample name for which report file should be generated.

  • congested

  • cpu_util—CPU utilization: milliseconds of time spent

  • disk_device_io—storage device I/O statistics

  • disk_io—operating system aggregate disk I/O: KB/sec

  • fan—fan speed

  • fs_mnt_bytes—filesystem usage: bytes

  • fs_mnt_inodes—filesystem usage: inodes

  • interface—network interface statistics

  • intf_util—network interface utilization: bytes

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

  • power—power supply usage

  • power-consumption

  • temperature—modules temperature

  • interface-ethernet—Ethernet counters statistics: counter units

  • interface-mlag-port-channel—MLAG counters statistics: counter units

  • interface-port-channel—LAG counters statistics: counter units

max-entries

Number of records
Range: 1-1000

Default

Default “interface” samples is 100 records

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats sample interface-ethernet max-entries 1000

Related Commands

show stats sample

Notes

  • Setting a new value will delete all sample history.

  • History does not persist after reboot.

stats clear-all

stats clear-all

Clears data for all samples, CHDs, and status for all alarms.

Syntax Description

N/A

Default

N/A

Configuration Mode

config

History

8.0.0100

Example

gateway (config) # stats clear-all

Related Commands

show stats sample

Notes

show stats alarm

show stats alarm [<alarm-id> [rate-limit]]

Displays status of all alarms or the specified alarm.

Syntax Description

alarm-id

Available values:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

rate-limit

Displays rate limit parameters

Default

N/A

Configuration Mode

Any command mode

History

8.0.0100

Example

gateway (config) # show stats alarm
Alarm cpu_util_indiv (Average CPU utilization too high): ok
Alarm disk_io (Operating System Disk I/O per second too high): (disabled)
Alarm fs_mnt (Free filesystem space too low): ok
Alarm intf_util (Network utilization too high): (disabled)
Alarm memory_pct_used (Too much memory in use): (disabled)
Alarm paging (Paging activity too high): ok
Alarm temperature (Temperature is too high): ok

Related Commands

stats alarm

Notes

show stats chd

show stats chd [<chd-id>]

Displays configuration of all statistics CHDs.

Syntax Description

chd-id

Available values:

  • cpu_util_indiv—average CPU utilization too high: percent utilization

  • disk_io—operating System Disk I/O per second too high: kilobytes per second

  • fs_mnt—free filesystem space too low: percent of disk space free

  • intf_util—network utilization too high: bytes per second

  • memory_pct_used—too much memory in use: percent of physical memory used

  • paging—paging activity too high: page faults

Default

N/A

Configuration Mode

Any command mode

History

8.0.0100

Example

gateway (config) # show stats chd disk_device_io_hour

CHD "disk_device_io_hour" (Storage device I/O read/write statistics for the last
hour: bytes):
Enabled: yes
Source dataset: sample "disk_device_io"
Computation basis: data points
Interval: 1 data point(s)
Range: 1 data point(s)

Related Commands

stats chd

Notes

show stats cpu

show stats cpu

Displays some basic stats about CPU utilization:

  • the current level

  • the peak over the past hour

  • the average over the past hour

Syntax Description

N/A

Default

N/A

Configuration Mode

Any command mode

History

8.0.0100

Example

gateway (config) # show stats cpu

CPU 0
Utilization: 6%
Peak Utilization Last Hour: 16% at 2012/02/28 08:47:32
Avg. Utilization Last Hour: 8%

Related Commands

Notes

show stats sample

show stats sample [<sample-id>]

Displays sampling interval for all samples, or the specified one.

Syntax Description

sample-id

Sample name for which report file should be generated.

  • congested

  • cpu_util—CPU utilization: milliseconds of time spent

  • disk_device_io—storage device I/O statistics

  • disk_io—operating system aggregate disk I/O: KB/sec

  • fan—fan speed

  • fs_mnt_bytes—filesystem usage: bytes

  • fs_mnt_inodes—filesystem usage: inodes

  • interface—network interface statistics

  • intf_util—network interface utilization: bytes

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

  • power—power supply usage

  • power-consumption

  • temperature—modules temperature

  • interface-ethernet—Ethernet counters statistics: counter units

  • interface-mlag-port-channel—MLAG counters statistics: counter units

  • interface-port-channel—LAG counters statistics: counter units

Default

N/A

Configuration Mode

Any command mode

History

8.0.0100

Example

gateway (config) # show stats sample fan
Sample "fan" (Fan speed):
Enabled: yes
Sampling interval: 1 minute 11 seconds

Related Commands

Notes

show stats sample data

show stats sample <sample-id> data [interface {ethernet | port-channel | mlag-port-channel} <device/port> [counter <counter-name>] ] [group name <group-name> [counter <counter-name>] ] [max-samples {<max-samples> | all}]

Displays history of counter values (i.e. collected information for a sample).

Syntax Description

sample-id

Sample name for which report file should be generated.

  • congested

  • cpu_util—CPU utilization: milliseconds of time spent

  • disk_device_io—storage device I/O statistics

  • disk_io—operating system aggregate disk I/O: KB/sec

  • fan—fan speed

  • fs_mnt_bytes—filesystem usage: bytes

  • fs_mnt_inodes—filesystem usage: inodes

  • interface—network interface statistics

  • intf_util—network interface utilization: bytes

  • memory—system memory utilization: bytes

  • paging—paging activity: page faults

  • power—power supply usage

  • power-consumption

  • temperature—modules temperature

  • interface-ethernet—Ethernet counters statistics: counter units

  • interface-mlag-port-channel—MLAG counters statistics: counter units

  • interface-port-channel—LAG counters statistics: counter units

interface

Allows limiting output to a particular interface’s counters

group

Allows limiting output to a particular group of counters

counter

Allows limiting output to a particular counter. This option is available only if the option interface or group is chosen.

max-samples

Allows choosing a number of counter records to display. Range: 1-1000 records. The “all” option is meant for all available records. By default, 20 counter records are displayed.

Default

N/A

Configuration Mode

Any command mode

History

8.0.0100

Example

gateway (config) # show stats sample interface-ethernet data interface ethernet 1/1 max-samples 1
Sampling data for Interface ethernet counters:
Eth1/1:
------------------------------------------------------------------
Name Timestamp Value
------------------------------------------------------------------
Rx_packets 2020/05/15 10:27:53 0
Rx_unicast_packets 2020/05/15 10:27:53 0
Rx_multicast_packets 2020/05/15 10:27:53 0
Rx_broadcast_packets 2020/05/15 10:27:53 0
Rx_bytes 2020/05/15 10:27:53 0
Rx_discard_packets 2020/05/15 10:27:53 0
Rx_error_packets 2020/05/15 10:27:53 0
Rx_fcs_errors 2020/05/15 10:27:53 0
Rx_undersize_packets 2020/05/15 10:27:53 0
Rx_oversize_packets 2020/05/15 10:27:53 0
Rx_pause_packets 2020/05/15 10:27:53 0
Rx_unknown_control_opcode 2020/05/15 10:27:53 0
Rx_symbol_errors 2020/05/15 10:27:53 0
Rx_packets_of_64_bytes 2020/05/15 10:27:53 0
Rx_packets_of_65-127_bytes 2020/05/15 10:27:53 0
Rx_packets_of_128-255_bytes 2020/05/15 10:27:53 0
Rx_packets_of_256-511_bytes 2020/05/15 10:27:53 0
Rx_packets_of_512-1023_bytes 2020/05/15 10:27:53 0
Rx_packets_of_1024-1518_bytes 2020/05/15 10:27:53 0
Rx_packets_Jumbo 2020/05/15 10:27:53 0
Tx_packets 2020/05/15 10:27:53 0
Tx_unicast_packets 2020/05/15 10:27:53 0
Tx_multicast_packets 2020/05/15 10:27:53 0
Tx_broadcast_packets 2020/05/15 10:27:53 0
Tx_bytes 2020/05/15 10:27:53 0
Tx_discard_packets 2020/05/15 10:27:53 0
Tx_error_packets 2020/05/15 10:27:53 0
Tx_hoq_discard_packets 2020/05/15 10:27:53 0
Tx_pause_packets 2020/05/15 10:27:53 0
Tx_pause_duration 2020/05/15 10:27:53 0
...

Related Commands

Notes

  • When there are more records in history than displayed, the output for a group ends with “...” (ellipses).

  • Filtering keyword depends on chosen <sample-id>. For convenience, “interface” samples such as “interface-ethernet”, “interface-port-channel” and “interface-mlag-port-channel” have interface related keywords for choosing a counters group.

  • Notice that this is a history of counters. Autocompletion and output can contain information for groups (interfaces) that is not present anymore in the system, and vice versa. If counters are not sampled, they will not appear in the output.

  • Output of collected information is implemented only for the following samples:

    • interface-port-channel

    • interface-ethernet

    • interface-mlag-port-channel

    • memory

    • paging

    • power

© Copyright 2023, NVIDIA. Last updated on May 23, 2023.