公开(公告)号：US10721137B2

公开(公告)日：2020-07-21

申请号：US15198300

申请日：2016-06-30

Applicant: Intel Corporation

Inventor： Mrittika Ganguli ,  Rahul Khanna ,  Ananth S. Narayan ,  Piyush Gupta

IPC: H04L12/24 ,  H04L29/08

Abstract: Systems described herein operate to improve network performance in a multi-tenant cloud computing environment. Systems can include communication circuitry and processing circuitry to generate a phase sequence matrix that indicates the identity and number of phases of a workload by measuring resources of the compute node during execution of the workload throughout a lifetime of the workload. The processing circuitry can generate a workload fingerprint that includes the phase sequence matrix and a phrase residency matrix. The phase residency matrix can indicate the fraction of execution time of the workload spent in each phase identified in the phase sequence matrix. A cloud controller can access the workload fingerprint for multiple workloads operating on multiple compute nodes in the cloud cluster to adjust workload allocations based at least on these workload fingerprints and on whether service level objectives (SLOs) are being met.

公开(公告)号：US10659386B2

公开(公告)日：2020-05-19

申请号：US15872928

申请日：2018-01-16

Applicant: Intel Corporation

Inventor： Subramony Sesha ,  Archana Patni ,  Ananth S. Narayan ,  Mrittika Ganguli

IPC: G06F15/173 ,  H04L12/911 ,  G06F9/50 ,  H04L29/08 ,  G06F9/455 ,  H04L12/26

Abstract: Technologies for contention-aware cloud compute scheduling include a number of compute nodes in a cloud computing cluster and a cloud controller. Each compute node collects performance data indicative of cache contention on the compute node, for example, cache misses per thousand instructions. Each compute node determines a contention score as a function of the performance data and stores the contention score in a cloud state database. In response to a request for a new virtual machine, the cloud controller receives contention scores for the compute nodes and selects a compute node based on the contention score. The cloud controller schedules the new virtual machine on the selected compute node. The contention score may include a contention metric and a contention score level indicative of the contention metric. The contention score level may be determined by comparing the contention metric to a number of thresholds. Other embodiments are described and claimed.

公开(公告)号：US09614779B2

公开(公告)日：2017-04-04

申请号：US14368349

申请日：2013-12-24

Applicant: Intel Corporation

Inventor： Subramony Sesha ,  Archana Patni ,  Ananth S. Narayan ,  Mrittika Mrittika Ganguli

IPC: G06F15/173 ,  H04L12/911 ,  H04L29/08

CPC classification number: H04L47/70 ,  G06F9/45504 ,  G06F9/5027 ,  G06F9/505 ,  H04L43/16 ,  H04L67/1095 ,  H04L67/1097 ,  H04L67/2842

Abstract: Technologies for contention-aware cloud compute scheduling include a number of compute nodes in a cloud computing cluster and a cloud controller. Each compute node collects performance data indicative of cache contention on the compute node, for example, cache misses per thousand instructions. Each compute node determines a contention score as a function of the performance data and stores the contention score in a cloud state database. In response to a request for a new virtual machine, the cloud controller receives contention scores for the compute nodes and selects a compute node based on the contention score. The cloud controller schedules the new virtual machine on the selected compute node. The contention score may include a contention metric and a contention score level indicative of the contention metric. The contention score level may be determined by comparing the contention metric to a number of thresholds. Other embodiments are described and claimed.

公开(公告)号：US20230327960A1

公开(公告)日：2023-10-12

申请号：US18210478

申请日：2023-06-15

Applicant: Intel Corporation

Inventor： Mrittika Ganguli ,  Muthuvel M. I ,  Ananth S. Narayan ,  Jaideep Moses ,  Andrew J. Herdrich ,  Rahul Khanna

IPC: H04L41/50 ,  H04L41/5025 ,  H04L41/14 ,  H04L41/5009

CPC classification number: H04L41/50 ,  H04L41/5009 ,  H04L41/145 ,  H04L41/5025

Abstract: In accordance with some embodiments, a cloud service provider may operate a data center in a way that dynamically reallocates resources across nodes within the data center based on both utilization and service level agreements. In other words, the allocation of resources may be adjusted dynamically based on current conditions. The current conditions in the data center may be a function of the nature of all the current workloads. Instead of simply managing the workloads in a way to increase overall execution efficiency, the data center instead may manage the workload to achieve quality of service requirements for particular workloads according to service level agreements.

公开(公告)号：US20220368645A1

公开(公告)日：2022-11-17

申请号：US17562664

申请日：2021-12-27

Applicant: Intel Corporation

Inventor： Subramony Sesha ,  Archana Patni ,  Ananth S. Narayan ,  Mrittika Ganguli

IPC: H04L47/70 ,  G06F9/50 ,  H04L67/568 ,  H04L67/1095 ,  G06F9/455 ,  H04L43/16 ,  H04L67/1097

Abstract: Technologies for contention-aware cloud compute scheduling include a number of compute nodes in a cloud computing cluster and a cloud controller. Each compute node collects performance data indicative of cache contention on the compute node, for example, cache misses per thousand instructions. Each compute node determines a contention score as a function of the performance data and stores the contention score in a cloud state database. In response to a request for a new virtual machine, the cloud controller receives contention scores for the compute nodes and selects a compute node based on the contention score. The cloud controller schedules the new virtual machine on the selected compute node. The contention score may include a contention metric and a contention score level indicative of the contention metric. The contention score level may be determined by comparing the contention metric to a number of thresholds. Other embodiments are described and claimed.

公开(公告)号：US11212235B2

公开(公告)日：2021-12-28

申请号：US16875597

申请日：2020-05-15

Applicant: Intel Corporation

Inventor： Subramony Sesha ,  Archana Patni ,  Ananth S. Narayan ,  Mrittika Ganguli

IPC: G06F15/173 ,  H04L12/911 ,  G06F9/50 ,  H04L29/08 ,  G06F9/455 ,  H04L12/26

Abstract: Technologies for contention-aware cloud compute scheduling include a number of compute nodes in a cloud computing cluster and a cloud controller. Each compute node collects performance data indicative of cache contention on the compute node, for example, cache misses per thousand instructions. Each compute node determines a contention score as a function of the performance data and stores the contention score in a cloud state database. In response to a request for a new virtual machine, the cloud controller receives contention scores for the compute nodes and selects a compute node based on the contention score. The cloud controller schedules the new virtual machine on the selected compute node. The contention score may include a contention metric and a contention score level indicative of the contention metric. The contention score level may be determined by comparing the contention metric to a number of thresholds. Other embodiments are described and claimed.

公开(公告)号：US20170237681A1

公开(公告)日：2017-08-17

申请号：US15388843

申请日：2016-12-22

Applicant: Intel Corporation

Inventor： Subramony Sesha ,  Archana Patni ,  Ananth S. Narayan ,  Mrittika Ganguli

IPC: H04L12/911 ,  G06F9/455 ,  G06F9/50 ,  H04L29/08 ,  H04L12/26

CPC classification number: H04L47/70 ,  G06F9/45504 ,  G06F9/5027 ,  G06F9/505 ,  H04L43/16 ,  H04L67/1095 ,  H04L67/1097 ,  H04L67/2842

Abstract: Technologies for contention-aware cloud compute scheduling include a number of compute nodes in a cloud computing cluster and a cloud controller. Each compute node collects performance data indicative of cache contention on the compute node, for example, cache misses per thousand instructions. Each compute node determines a contention score as a function of the performance data and stores the contention score in a cloud state database. In response to a request for a new virtual machine, the cloud controller receives contention scores for the compute nodes and selects a compute node based on the contention score. The cloud controller schedules the new virtual machine on the selected compute node. The contention score may include a contention metric and a contention score level indicative of the contention metric. The contention score level may be determined by comparing the contention metric to a number of thresholds. Other embodiments are described and claimed.

公开(公告)号：US11722382B2

公开(公告)日：2023-08-08

申请号：US17494569

申请日：2021-10-05

Applicant: Intel Corporation

Inventor： Mrittika Ganguli ,  Muthuvel M. I ,  Ananth S. Narayan ,  Jaideep Moses ,  Andrew J. Herdrich ,  Rahul Khanna

IPC: H04L12/24 ,  H04L41/50 ,  H04L41/5025 ,  H04L41/14 ,  H04L41/5009

CPC classification number: H04L41/50 ,  H04L41/145 ,  H04L41/5009 ,  H04L41/5025

Abstract: In accordance with some embodiments, a cloud service provider may operate a data center in a way that dynamically reallocates resources across nodes within the data center based on both utilization and service level agreements. In other words, the allocation of resources may be adjusted dynamically based on current conditions. The current conditions in the data center may be a function of the nature of all the current workloads. Instead of simply managing the workloads in a way to increase overall execution efficiency, the data center instead may manage the workload to achieve quality of service requirements for particular workloads according to service level agreements.

公开(公告)号：US10747640B2

公开(公告)日：2020-08-18

申请号：US15635270

申请日：2017-06-28

Applicant: INTEL CORPORATION

Inventor： Ananth S. Narayan ,  Lukasz Grzymkowski ,  Mrittika Ganguli ,  Mariusz Oriol

IPC: G06F11/34 ,  G06F9/38 ,  H04Q9/04 ,  H03K19/0175 ,  G06F9/50 ,  H04Q9/00 ,  G06F30/327 ,  G06F9/48

Abstract: Techniques and apparatus for managing a distributed computing environment using event digests are described. In one embodiment, for example, an apparatus may include at least one memory, and logic for a system manager, at least a portion of the logic comprised in hardware coupled to the at least one memory, the logic to determine a workload to schedule, access an event digest associated with a plurality of compute hosts, the event digest comprising event digest values determined using out-of-band information, determine metrics from the event digest, generate at least one host weight for at least a portion of the plurality of compute hosts based on the metrics, identify at least one candidate host from the portion of the plurality of compute hosts based on the at least one host weight, and schedule the workload on the at least one candidate host. Other embodiments are described and claimed.

公开(公告)号：US20190065261A1

公开(公告)日：2019-02-28

申请号：US15859366

申请日：2017-12-30

Applicant: Intel Corporation

Inventor： Ananth S. Narayan ,  Sagar V. Dalvi ,  Mrittika Ganguli ,  Sergiu D. Ghetie

IPC: G06F9/50 ,  G06F11/34 ,  G06F11/30 ,  G06F17/30

Abstract: Technologies for providing in-processor workload phase detection include a sled having a compute engine, which itself includes a performance monitor unit. The compute engine obtains telemetry data from the performance monitor unit. The performance monitor unit produces telemetry data indicative of performance metrics of the sled during execution of one or more workloads. The telemetry data is indicative of a resource utilization and workload performance by the sled as the workloads are executed. The compute engine determines, from a lookup table indicative of resource utilization phases, a resource utilization phase based on the obtained telemetry data. A workload fingerprint is updated based on the determined resource utilization phase, and the workload fingerprint is output. Other embodiments are also described and claimed.