Out-of-band platform tuning and configuration

Abstract

Devices and techniques for out-of-band platform tuning and configuration are described herein. A device can include a telemetry interface to a telemetry collection system and a network interface to network adapter hardware. The device can receive platform telemetry metrics from the telemetry collection system, and network adapter silicon hardware statistics over the network interface, to gather collected statistics. The device can apply a heuristic algorithm using the collected statistics to determine processing core workloads generated by operation of a plurality of software systems communicatively coupled to the device. The device can provide a reconfiguration message to instruct at least one software system to switch operations to a different processing core, responsive to detecting an overload state on at least one processing core, based on the processing core workloads. Other embodiments are also described.

1 . (canceled) 
     
     
         2 . An orchestration controller for a computer system having a multi-core computing platform architecture and a plurality of network adapters to provide in-band resources for facilitating in-band data flow for at least one software system, the orchestration controller comprising:
 a network adapter out-of-band (OOB) interface to collect network adapter hardware operational data via network adapter OOB access to network adapter hardware of the plurality of network adaptors;   wherein the network adapter OOB access is separate from the in-band resources.   
     
     
         3 . The orchestration controller of  claim 2 , wherein the network adapter hardware operational data includes silicon hardware statistics of the network adapter hardware. 
     
     
         4 . The orchestration controller of  claim 2 , further comprising:
 a platform out-of-band (OOB) interface to collect platform operational data via platform OOB access to processing cores of the multi-core computing platform, wherein the platform OOB access is separate from the in-band resources.   
     
     
         5 . The orchestration controller of  claim 4 , wherein the platform operational data includes platform telemetry metrics of a plurality of the processing cores. 
     
     
         6 . The orchestration controller of  claim 5 , wherein the platform telemetry metrics include processing core workloads of the plurality of the processing cores. 
     
     
         7 . The orchestration controller of  claim 5 , wherein the platform telemetry metrics include at least one metric selected from a group consisting of: processing core data, chipset data, memory element performance data, data received from an encryption unit, data received from a compression unit, storage data, virtual switch (vSwitch) data, or any combination thereof. 
     
     
         8 . The orchestration controller of  claim 5 , wherein the platform telemetry metrics include network interface card (NIC) telemetry data received over a NIC connection, including an indication of packets per second received at the NIC, average packet size received at the NIC, or some combination thereof. 
     
     
         9 . The orchestration controller of  claim 5 , wherein the platform telemetry metrics include platform quality of service (PQoS) metrics. 
     
     
         10 . The orchestration controller of  claim 4 , further comprising processing circuitry configured to:
 receive the platform operational data via the platform OOB interface, and receive the network adapter hardware operational data via the network adapter OOB interface to gather collected statistics,   determine processing core workloads generated by operation the at least one software system executed by the computer system, and   provide a reconfiguration message to instruct the at least one software system to shift operations between processing cores, responsive to the processing core workloads.   
     
     
         11 . The orchestration controller of  claim 10 , wherein the reconfiguration message is to instruct the at least one software system to switch certain operations from a first processing core to a second processing core. 
     
     
         12 . The orchestration controller of  claim 10 , wherein the processing circuitry is further configured to:
 determine whether service level agreement (SLA) criteria have been met based on the processing core workloads; and   report a SLA violation to a datacenter management entity if the SLA criteria have not been met.   
     
     
         13 . The orchestration controller of  claim 10 , wherein the processing circuitry is further configured to:
 instruct a set of at least two processing cores to enter an offline state;   provide instructions for performing testing on each of the set of at least two processing cores after a respective one of the set of at least two processing cores has entered the offline state; and   rank performance of at least two processing cores, based on respective performance of those processing cores during the testing, to produce a ranked set.   
     
     
         14 . The orchestration controller of  claim 13 , wherein the processing circuitry is further configured to:
 provide instructions for steering incoming NIC traffic to a processing core of the ranked set based on priority level of the incoming NIC traffic.   
     
     
         15 . The orchestration controller of  claim 13 , wherein the processing circuitry is further configured to:
 receive a configuration state from a remote entity, the configuration state including at least one processing core identifier and at least one configuration parameter corresponding to the at least one processing core identifier;   provide, to the remote entity, measured performance of at least one processing core identified by the at least one processing core identifier based on the testing; and   receive reconfiguration information from the remote entity in response to the measured performance.   
     
     
         16 . The orchestration controller of  claim 10 , wherein the processing circuitry is further configured to:
 in response to receipt of performance monitoring event information corresponding to a parameter of interest, detect application performance to generate a performance measure associating application performance to the parameter of interest;   generate a sensitivity relation, based on the performance measure, to determine sensitivity of application performance to the parameter of interest; and   provide the sensitivity relation as an input to a reconfiguration decision algorithm that produces the reconfiguration message.   
     
     
         17 . An automated method for managing resources in a computer system having a multi-core computing platform architecture and a plurality of network adapters, the method comprising:
 communicating in-band data flow for at least one software system via in-band resources of the computer system;   collecting network adapter hardware operational data via a network adapter out-of-band (OOB) access to network adapter hardware of the plurality of network adaptors;   wherein the network adapter OOB access is separate from the in-band resources.   
     
     
         18 . The method of  claim 17 , wherein collecting the network adapter hardware operational data includes collecting silicon hardware statistics of the network adapter hardware. 
     
     
         19 . The method of  claim 17 , further comprising:
 collecting platform operational data via platform OOB access to processing cores of the multi-core computing platform, wherein the platform OOB access is separate from the in-band resources.   
     
     
         20 . The method of  claim 19 , further comprising:
 receiving the platform operational data via the platform OOB access;   receiving the network adapter hardware operational data via the network adapter OOB access to gather collected statistics;   determining processing core workloads generated by operation the at least one software system executed by the computer system, and   providing a reconfiguration message to instruct the at least one software system to shift operations between processing cores, responsive to the processing core workloads.   
     
     
         21 . The method of  claim 20 , further comprising:
 detecting any presence of an overload state on at least one of the processing cores, based on the processing core workloads.   
     
     
         22 . The method of  claim 20 , further comprising:
 determining whether service level agreement (SLA) criteria have been met based on the processing core workloads; and   reporting a SLA violation to a datacenter management entity if the SLA criteria have not been met.   
     
     
         23 . The method of  claim 20 , further comprising:
 instructing a set of at least two processing cores to enter an offline state;   providing instructions for performing testing on each of the set of at least two processing cores after a respective one of the set of at least two processing cores has entered the offline state; and   ranking performance of at least two processing cores, based on respective performance of those processing cores during the testing, to produce a ranked set.   
     
     
         24 . The method of  claim 23 , further comprising:
 providing instructions for steering incoming NIC traffic to a processing core of the ranked set based on priority level of the incoming NIC traffic.   
     
     
         25 . The method of  claim 23 , further comprising:
 receiving a configuration state from a remote entity, the configuration state including at least one processing core identifier and at least one configuration parameter corresponding to the at least one processing core identifier;   providing, to the remote entity, measured performance of at least one processing core identified by the at least one processing core identifier based on the testing; and   receiving reconfiguration information from the remote entity in response to the measured performance.   
     
     
         26 . The method of  claim 20 , further comprising:
 in response to receipt of performance monitoring event information corresponding to a parameter of interest, detecting application performance to generate a performance measure associating application performance to the parameter of interest;   generating a sensitivity relation, based on the performance measure, to determine sensitivity of application performance to the parameter of interest; and   providing the sensitivity relation as an input to a reconfiguration decision algorithm that produces the reconfiguration message.