Cloud service monitoring system

Patent metadata

Field	Value
Publication number	US-9985847-B2
Application number	US-201314079383-A
Country	US
Kind code	B2
Filing date	Nov 13, 2013
Priority date	Sep 7, 2011
Publication date	May 29, 2018
Grant date	May 29, 2018

Abstract

A cloud services monitoring system (CSMS) helps a consumer of cloud computing resources determine the state of the cloud computing resources available to the consumer. The CSMS is particularly effective for cloud models where the consumer of the cloud computing resources is not the same as the provider of the cloud resource. The CSMS also provides mechanisms for adjusting operations actions to preserve service guarantees for the consumer. In that regard, the CSMS dynamically adjusts operations actions in response to the state of the cloud, rather than executing fixed predetermined operations actions that are not tailored to the monitored conditions.

First claim

What is claimed:

1. A computer-implemented method for cloud services monitoring (CSM), comprising:
identifying, with a processor, a service level agreement (SLA) rule to execute to meet a SLA requirement included in a SLA for a user of cloud resources provided by a cloud computing environment comprising multiple computing zones, the SLA rule being accessed by the processor using a network, where the SLA rule is further configured to comprise: composite service rule scenarios assigning executable compute instances for each of the multiple computing zones, wherein execution of a first compute instance assigned to a first computing zone included in the multiple computing zones expends cloud resources of the first computing zone;
observing, using the processor connected to the network, the cloud computing environment;
collecting, using the processor connected to the network, state information for the cloud computing environment;
detecting, using the processor, an occurrence of a predetermined event within the cloud computing environment;
identifying, using the processor, a mathematical model from a plurality of available mathematical models based on a number of detected occurrences of the predetermined event, wherein the plurality of available mathematical models includes at least a linear regression model and a linear least squared estimate model;
applying, using the processor, the identified mathematical model to the collected state information;
determining, using the processor, whether the state information indicates that the cloud computing environment is at risk of not meeting the SLA requirement based on the application of the identified mathematical model to the collected state information;
based on a determination that the cloud computing environment is at risk of not meeting the SLA requirement, adjusting the SLA rule to operate a different number of compute instances in the first computing zone that improves the probability that the cloud computing environment, including the first computing zone, will meet the SLA requirement; and
executing the different number of compute instances in the first computing zone to meet the SLA requirement.

2. The method of claim 1 , wherein the multiple computing zones further includes a second computing zone, and the multiple computing zones are organized with respect to respective achievable SLAs; and
wherein the achievable SLAs for the first computing zone and the second computing zone are different.

3. The method of claim 1 , where observing comprises:
obtaining a measurement of network delay, or bandwidth of the cloud computing environment, or both, for the multiple computing zones including the first computing zone and a second computing zone by:
sending via the network a file transfer service request to the cloud computing environment for a file of known size to transfer from a first CSM instance to a second CSM instance located at known locations;
receiving via the network a result of the file transfer service request; and
calculating the measurement of network delay, or the bandwidth of the cloud computing environment, or both based on the result of the file transfer service request.

4. The method of claim 1 , where observing comprises:
obtaining a measurement of time-to-start of a computing instance that executes in the cloud computing environment, or spot price for any of the cloud resources used to meet the SLA, or both, for the multiple computing zones including the first computing zone and a second computing zone.

5. The method of claim 4 , where adjusting the SLA rule comprises:
changing the SLA rule to have a different time-to-start for a computing instance that executes in the cloud computing environment.

6. The method of claim 1 , where adjusting the SLA rule comprises:
changing the SLA rule to have a different time-to-request a selected cloud resource among the cloud resources by communicating the SLA rule change to an orchestration tool connected, via the network, to the cloud computing environment; and
wherein the orchestration tool implements the SLA rule change.

7. The method of claim 1 , where adjusting the SLA rule comprises:
changing the SLA rule to direct a portion of cloud computing workload for the user to a different cloud computing environment.

8. A product for cloud services monitoring (CSM), comprising:
a computer readable memory with processor executable instructions stored thereon, wherein the instructions when executed by the processor cause the processor to:
identify a service level agreement (SLA) rule configured to execute in order to meet a SLA requirement for a user of cloud resources provided by a cloud computing environment comprising multiple computing zones, where the SLA rule is further configured to comprise: composite service rule scenarios assigning executable compute instances for each of the multiple computing zones, wherein execution of a first compute instance assigned to a first computing zone included in the multiple computing zones expends cloud resources of the first computing zone;
observe the cloud computing environment;
collect state information for the cloud computing environment;
detect an occurrence of a predetermined event within the cloud computing environment;
identify, from a plurality of available mathematical models including at least a linear regression model and a linear least squared estimate model, a mathematical model based on a number of detected occurrences of the predetermined event;
apply the identified mathematical model to the collected state information;
determine whether the state information indicates that the cloud computing environment is at risk of not meeting the SLA requirement based on the application of the identified mathematical model to the collected state information;
adjust the SLA rule to operate a different number of compute instances in the first computing zone that will improve the probability that the cloud computing environment, including the first computing zone, will meet the SLA requirement; and
execute the different number of compute instances in the first computing zone to meet the SLA requirement.

9. The product of claim 8 , wherein the instructions further cause the processor to:
obtain passive state information directly provided by the cloud computing environment; and
obtain detected state information by probing the cloud computing environment.

10. The product of claim 8 , wherein the instructions further cause the processor to:
obtain a measurement of network delay, or bandwidth of the cloud computing environment, or both for the multiple computing zones including the first computing zone and a second computing zone by:
sending via the network a file transfer service request to the cloud computing environment for a file of known size to transfer from a first CSM instance to a second CSM instance located at known locations;
receiving via the network a result of the file transfer service request; and
calculating the measurement of network delay, or the bandwidth of the cloud computing environment, or both based on the result of the file transfer service request.

11. The product of claim 8 , wherein the instructions further cause the processor to:
obtain a measurement of time-to-start of a computing instance that executes in the cloud computing environment, or spot price for any of the cloud resources used to meet the SLA, or both, for the multiple computing zones including the first computing zone and a second computing zone.

12. The product of claim 11 , wherein the instructions further cause the processor to:
change the SLA rule to have a different time-to-request a selected clou

Assignees

• Accenture Global Services Ltd

Inventors

• Tung Teresa
• Badruddoja Farzana
• Kang Jaeyoung C

Classifications

• G06F11/30

  Monitoring · CPC title

• G06F11/3495

  for systems · CPC title

• H04L41/5003 Primary

  Managing SLA; Interaction between SLA and QoS · CPC title

• G06F2209/503

  Resource availability · CPC title

• G06F2201/815

  Virtual · CPC title