In this section:

Overview

This section describes the following aspects of creating a GPU ISBC on KVM:

  • Host configurations 
    -  Enabling GPU passthrough and CPU pinning on KVM host.

  • VM configurations 
    - Launching the KVM instance 
    - Associating the GPU PCIe device to the instance
    - Enabling CPU pinning

  • Installation procedure
    - OS and SBC Application installation

  • SBC configurations
    - Applying custom GPU traffic profile.

Host Configurations

This section describes the changes required on the KVM host to enable:

    • GPU passthrough and 
    • CPU pinning


OS Configuration for KVM host with GPU Device

Step
Action
Notes
1

Edit /etc/sysconfig/grub to ensure that the following parameters are populated:

intel_iommu=on iommu=pt rdblacklist=nouveau
  • Enables kernel support for PCIe passthrough.
  • Blacklists opensource NVIDIA driver (nouveau) from loading on the host.
2

Update grub using the following command:

grub2-mkconfig -o /etc/grub2.cfg
3

Edit /etc/systemd/system.conf to the CPUAffinity parameter is uncommented

and it is populated with the CPUs that needs to be left to the host processes.

Isolates the remaining CPUs from the host.

4

Create /etc/modprobe.d/nouveau.conf file with the following contents:

blacklist nouveau
Blacklists opensource NVIDIA driver (nouveau) from loading on the host.
5Reboot the KVM host.

 



VM Configurations 

This section covers following aspects:

  1. Creation of KVM instance.
  2. Associating the GPU PCIe device to the KVM instance.
  3. Apply CPU pinning configurations to the KVM instance.

Steps in detail:

    1. Select Add Hardware option as shown below:

      Add Hardware


    2. Select PCI Host Device option as shown below:

      PCI Host Device


    3. From the list of PCI devices, scroll and select the GPU PCIe device/s to be allocated to the instance as shown below:

      Select the GPU PCIe Device/s


    4. Select Finish to allocate the selected PCIe device.
    5. Start the instance.

SBC Installation Procedure

SBC Configuration

After installing the OS and the SBC, the application will be up with the default traffic profile being activated.

To apply the custom GPU profile, execute the following CLI command by entering the CLI mode using admin user (enter into the configuration mode):


  • Create a codecMix profile with the codecs to be configured for GPU transcoding:

    % set system sweCodecMixProfile <profile name> <codec> <ptime value: 10ms, 20ms, 30ms, 40ms, 60ms, 80ms, 100ms> percentage <percentage value>



    Parameter
    Length/Range
    Description
    <codec>N/A

    Enter the codec used by this Codec Mix Profile. Select one of the currently supported codecs:

    • Codecs supported on both GPU and CPU: AMR-NB, AMR-WB, EVRC, EVRCB, G729, G722, EVS, OPUS 
    • CPU only codecs :  G723, G726, G7221, ILBC, SILK_8, SILK_16, G7112G711, T38,G711

    Note:

    - For more information on "GPU + CPU codecs", see the section Dynamic Codec Allocation.

    - You can provision CPU codecs in the codec profile and associate it with the GPU traffic profile; however, you must provision at least one GPU codec in the sweCodecMixProfile.

    <ptime value>N/A

    Select a packetization time value representing 10 ms, 20 ms, 30 ms, 40 ms, 60 ms:

    • p10
    • p20
    • p30
    • p40
    • p60
    • p80
    • p100

    Note: For SBC SWe Cloud 8.1, ptime values p80 and p100 are not tested.

    percentage0.00-100.00<% value> – The percentage distribution allocated for the codec mix. The sum of all percentage entries of all columns in any row of the transcoding profile table must equal 100.

    sweCodecMixProfile1-40 characters

    <profile name> – Enter a unique SWe Codec Mix Profile name or one of the default profiles:

    The SWe Codec Mix Profile is attachable to transcodingCodecProfile objects of the SWe Traffic Profile.

    For example, to configure a codec profile for AMRWB-G711u transcoding, run the following command:

    % set system sweCodecMixProfile customCodecMix amrwb p20 percentage 50
% set system sweCodecMixProfile customCodecMix g711  p20 percentage 50
% commit


     

  1. Create a traffic profile. Set the transcodingCodecProfile with the codecMix profile created in step 1.

    % set system sweTrafficProfiles <profile name> isAccess <value> callHoldTime <duration> passthroughCodecProfile <profile name> transcodePercent <percentage> transcodingCodecProfile <profile name> useGPUForTranscoding <value>
    Parameter
    Length/Range
    Description
    sweTrafficProfiles1-40 characters

    <profile name> – Enter a unique SWe Traffic Profile name.

    NOTE: To create additional profiles, delete any inactive custom profiles.

    isAccessN/A

    Set this flag to "true" to specify whether the deployment uses an access scenario.

    • false (default)
    • true

    NOTE: When set to 'true', internalRefreshTimer, registrationRefreshInterval and bhcaPerSubscriber parameters are available for configuration.

    bhcaPerSubscriber0-5

    <# attempts> (default = 1) – Indicates busy hour call attempts (BHCA) per subscriber.

    NOTE: This parameter is available when accessScenario = "true".

    callHoldTime

    		10-10800<# seconds> – Enter the average call hold time, in seconds, of the call load for this profile. (default = 90)
    cryptoPercent0-100

    <% value> (default = 0) – The percentage of media sessions (including both transcoding and passthrough) requiring cryptographic treatment.The value is 50 when there is SRTP<->RTP interworking on all calls.

    directMediaPercent0-100

    <% value> (default 0) – The call load percentage for direct media.

    NOTE: The combined total percentage of directMediaPercent and transcodePercent should not exceed 100%.

    passthroughCodecProfileN/A

    The name of the codec mix to associate with the Passthrough Codec Profile.

    • G711_20ms
    processorCapabilityIndexOverrideN/A

    Use this flag to enable/disable index overriding of the default CPU performance computation by the SBC SWe. When set to 'true', the computed indices (which are calculated during system boot-up) are ignored, and the value provided in processorCapabilityIndexOverrideValue attribute is used for all estimations.

    • false (default)
    • true

    NOTE: Since use cases for overriding the default computed indices are rare, it is recommended to not set the value of processorCapabilityIndexOverride to "true" to avoid inaccurate session numbers and vCPU computations.

    processorCapabilityIndexOverrideValue0.2-10

    Use this parameter to specify the computational value to use to override the default computed indices. (default = 1)

    NOTE: This parameter is available when processorCapabilityIndexOverride is set to "true". The value given should be in 0.2 to 10.

    internalRefreshTimer15-86400

    <# seconds> (default = 1800) – Use this parameter to specify the internal registration timer, in seconds.

    NOTE: The parameter is available only when isAccess is set to "true".

    externalRefreshTimer15–86400

    <# seconds> (default = 1800) – Use this parameter to specify the external registration timer, in seconds.

    NOTE: The parameter is available only when isAccess is set to "true".

    transcodePercent0-100

    <% value> (default = 0) – Use this parameter to specify the percentage of call load to use for transcoded calls.

    NOTE: The combined total percentage of directMediaPercent and transcodePercent should not exceed 100%.

    transcodingCodecProfileN/A

    The name of the codec mix to associate with the Transcoding Codec Profile.

    tonesPercent0-100<% value> (default = 0) – Use this parameter to specify the percentage of legs to use for tones treatment.
    useGPUForTranscodingN/A

    Set this flag to "true" to specify whether the deployment uses GPU for transcoding.

    • false (default)
    • true


    For example: To create a custom GPU profile with the name custom_gpu_profile using the codecMixProfile created in step 1 for trascoding profile:

    % set system sweTrafficProfiles custom_gpu_profile callHoldTime 100 transcodePercent 50 passthroughCodecProfile G711_20ms transcodingCodecProfile customCodecMix useGPUForTranscoding true
% commit



  2. Activate the traffic profile created in Step 2.

     % set system sweActiveProfile name <profile name>

    Where the profile name is the custom traffic profile name given in step 2.

    For example, to activate the custom GPU traffic profile created in step 2, run the following command:

    % set system sweActiveProfile name custom_gpu_profile
% commit

    The instance will reboot and comes up in the activated traffic profile.

  3. To check which profile is activated, run the following command:

    > show table system sweActiveProfile


    Example: 

    admin@vsbc1> show table system sweActiveProfile
name custom_gpu_profile;
stateChangeTime 2018-12-09T13:25:43-00:00;
[ok][2018-12-11 01:29:03]
admin@vsbc1>


 

Dynamic Codec Allocation

The SBC dynamically manages GPU and CPU transcoding resources so they respond to various codec combinations without impacting service. The SBC leverages both GPU and CPU transcoding resources to support an incoming codec request. If GPU resources are available, the SBC utilizes these GPU resources to support GPU-supported codecs. If GPU resources are not available, the SBC utilizes CPU resources (if available) to support GPU-supported codecs. The SBC utilizes CPU resources (if available) to support non-GPU supported codecs. For information about GPU- and CPU-supported codecs, refer to Supported Codecs and Transcoding.

Note

The percentage value for G7112G711 is used for estimating transcode and bandwidth cost.

The percentage value for G711 is not used for estimating transcode cost, but is used for bandwidth calculation of PXPAD scenarios.

The percentage value for G711 cannot be greater than the percentage value of non-G711 codecs.

The sum of all codec percentages is 100.

DSP-based Tone detection is supported only on GPU-ISBC profile.

Note

With dynamic GPU resource management, the sweTrafficProfiles does not enforce any hard coded per-codec limit. You must still configure the codec percentage in sweCodecMixProfile since this configuration acts as an input for the functional partitioning of VCPUs of the instance.

This feature does not affect the codec channel capacities in a non-GPU accelerated SWe (CPU-based SWe).

This feature supports the standard LSWU upgrade operation of the supported SBC releases.

The SBC NRM congestion monitoring mechanism monitors all UXPAD processes for congestion.

Note

A sub-optimal resource utilization exists during the transient transcoding call mix scenarios.

Best Practice

Ribbon recommends monitoring of the status of codec channel capacities for codecs provisioned on GPU (on a per GPU device basis), as well as on CPU.

GPU Transcode Status

Shows codec channel capacities on a per GPU device basis.

Note

The gpuTranscodeStatus represents only available GPU resources. If the gpuTranscodeStatus displays higher numbers than the dspStatus, the system is CPU-limited.

Command Syntax

show table system gpuTranscodeStatus


Command Parameters


gpuTranscodeStatus Descriptions

ParameterStatus Names and Descriptions

gpuTranscodeStatus

<system name> – The SBC system name.

  • amrNbTotal – Total AMR–NB resource capacity on this server.

  • amrNbUtilization – Percentage utilization of AMR–NB resources on this server.

  • amrWbTotal – Total AMR–WB resource capacity on this server.

  • amrWbUtilization – Percentage utilization of AMR–WB resources on this server.
  • evrc0Total – Total EVRC0 resource capacity on this server.
  • evrc0Utilization – Percentage utilization of EVRC0 resources on this server.
  • evrcb0Total – Total EVRCB0 resource capacity on this server.
  • evrcb0Utilization – Percentage utilization of EVRCB0 resources on this server.
  • gpuAllocation – Displays the overall GPU occupancy in a percentage.
  • gpuNumber - Numeric identifier for the GPU device.
  • g722Total – Total G.722 resource capacity on this server.
  • g722Utilization – Percentage utilization of G.722 resources on this server.
  • g729AbTotal – Total G729A+B resource capacity on this server.
  • g729AbUtilization – Percentage utilization of G729A+B resources on this server.

  • opusTotal – Total OPUS resource capacity on this server. The SBC does not support this parameter.

  • opusUtilization – Percentage utilization of OPUS resources on this server. The SBC does not support this parameter.


CPU Transcode Status

The cpuTranscodeStatus captures the information on spilled over channels due to the exhaustion of GPU resources. GPU codecs appear on the CPU when the GPU resources are exhausted and cannot accommodate GPU-supported codec channels, so the codec channels spill onto the CPU.

Command Syntax

show table system cpuTranscodeStatus

Command Parameters


cpuTranscodeStatus Descriptions

ParameterStatus Names and Descriptions

cpuTranscodeStatus

<system name> – The SBC system name.

  • amrNbUsed - Number of AMR channels spilled over onto CPU.

  • amrWbUsed - Number of AMR-WB channels spilled over onto CPU.

  • evrc0Used - Number of EVRC channels spilled over onto CPU.

  • evrcb0Used - Number of EVRCB channels spilled over onto CPU.

  • evsUsed - Number of EVS channels spilled over onto CPU.
  • opusUsed - Number of OPUS channels spilled over onto CPU. 
  • g722Used - Number of G.722 channels spilled over onto CPU.
  • g729AbUsed - Number of G.729AB channels spilled over onto CPU.

To display the status using the EMA, refer to: