Managing workloads and hardware resources in a cloud resource

• US 8,959,217 B2
• Filed: 01/29/2010
• Issued: 02/17/2015
• Est. Priority Date: 01/15/2010
• Status: Active Grant

• Current Assignee: Joyent, Inc. (Samsung Electronics Co. Ltd.)
• Original Assignee: Joyent, Inc. (Samsung Electronics Co. Ltd.)
• Inventors: Mayo, Mark G, Duncan, James, Hoffman, Jason A, Young, David P, Candel, Pedro Palazn
• Primary Examiner(s): CHRISTENSEN, SCOTT B
• Application Number:
  US12/696,334
• Publication Number:
  US 20110179162A1
• Time in Patent Office: 1,845 Days
• Field of Search: 709/224
• US Class Current: 709/226
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455
• CPC Class Codes: H04L 12/1822 Conducting the conference, e.g. admission, detection, selection or grouping of participants, correlating users to one or more conference sessions, prioritising transmission
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04L 69/329 in the application layer [OSI layer 7]
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

Systems and methods to manage workloads and hardware resources in a data center or cloud. In one embodiment, a method includes a data center having a plurality of servers in a network. The data center provides a virtual machine for each of a plurality of users, each virtual machine to use a portion of hardware resources of the data center. The hardware resources include storage and processing resources distributed onto each of the plurality of servers. The method further includes sending messages amongst the servers, some of the messages being sent from a server including status information regarding a hardware resource utilization status of that server. The method further includes detecting a request from the virtual machine to handle a workload requiring increased use of the hardware resources, and provisioning the servers to temporarily allocate additional resources to the virtual machine, wherein the provisioning is based on status information provided by one or more of the messages.

Claims

1. A method comprising:

monitoring, by a cloud control system, a plurality of workloads being handled by a plurality of servers in a cloud, each workload being initiated based on a request by a virtual machine, each server providing one or more shared hardware resources, and each workload using one or more dedicated hardware resources, wherein each workload is allocated to the one or more hardware resources;

detecting, by the cloud control system, a high demand placed by a first workload of the plurality of workloads upon a first server of the plurality of servers, wherein the high demand is greater than an average demand of other workloads being handled by the first server, and wherein the first workload is using a first shared hardware resource of the first server; and

based on detecting the high demand, moving the first workload of the first server from the first shared hardware resource to at least one dedicated hardware resource of the first server,

wherein the at least one dedicated hardware resource of the first server is only used by resources that are moved responsive to the detecting of high demand.

2. The method of claim 1, wherein each respective server of the plurality of servers comprises:

at least one physical processor providing a plurality of logical processors, the logical processors including a shared logical processor and a dedicated logical processor, and the dedicated logical processor configured to operate as the dedicated hardware resource;

system memory coupled for access by each of the logical processors, the system memory comprising at least four gigabytes of memory for each logical processor; and

a storage device configured to operate as the dedicated hardware resource.

3. The method of claim 1, wherein each respective server of the plurality of servers comprises:

at least one physical processor providing four or more processing cores, each core using simultaneous multithreading;

system memory coupled for access by each of the processing cores, the system memory comprising at least eight gigabytes of memory for each processing core;

at least one hard disk drive (HDD) having an HDD total storage capacity, at least a portion of the HDD configured to provide a shared storage resource; and

at least one solid state drive (SSD) having an SSD total storage capacity, at least a portion of the SSD configured to provide a dedicated storage resource, and wherein the SSD capacity is at least 35 percent of the HDD capacity;

wherein a total memory size of the system memory is at least ten percent of the combined HDD capacity and SSD capacity.

4. The method of claim 1, wherein each respective server of the plurality of servers comprises:

at least one physical processor providing four or more processing cores;

system memory coupled for access by each of the processing cores, the system memory comprising at least four gigabytes of memory for each processing core; and

at least one solid state drive (SSD), at least a portion of the SSD configured to provide the dedicated storage resource, and wherein each SSD has a data input/output rate of at least ten megabytes per second for each gigabyte of system memory.

5. The method of claim 2, wherein:

each respective server further comprises a hard disk drive to provide its respective one or more shared hardware resources; and

the storage device is a solid state drive (SSD), the SSD including at least five logical drives configured to shift back-and-forth between operation as a shared resource or as a dedicated resource as requested during operation of the cloud.

6. The method of claim 2, wherein the storage device has a data input/output rate of at least ten megabytes per second for each gigabyte of system memory.

7. The method of claim 6, wherein the storage device is a solid state drive (SSD), and each server further comprises at least one hard disk drive coupled for read/write access by at least one of the logical processors.

8. The method of claim 1, wherein:

the cloud control system is in communication with the plurality of servers;

monitoring the plurality of workloads comprises executing an agent on each of the plurality of servers, each agent including a process that monitors utilization of system resources on the respective server and that reports server utilization data to the cloud control system; and

moving the first workload is performed at least in part based on an economic model stored at the cloud control system and based on an analysis of server resource data and historical user data by the cloud control system, the economic model associating a cost with at least one hardware resource.

9. The method of claim 1, wherein:

the cloud control system is one of the plurality of servers; and

monitoring the plurality of workloads comprises executing an agent on each of the plurality of servers, each agent including a process that monitors utilization of system resources on the respective server.

10. The method of claim 1, wherein the high demand is characterized by use of a data input/output rate for a storage device of the first server, or by a processing load on a logical processor of the first server.

11. The method of claim 1, wherein each of the shared hardware resources is a processor or a storage device, and the high demand is characterized by a processing load on the processor or by use of a data input/output rate for the storage device.

12. The method of claim 11, wherein the first shared hardware resource is a shared core of a processor of the first server, and the dedicated hardware resource is a dedicated core of a processor of the first server.

13. The method of claim 12, wherein the processor is a single physical processor comprising the shared core and the dedicated core.

14. The method of claim 11, wherein the first shared hardware resource is a shared storage area of a storage device of the first server, and the dedicated hardware resource is a dedicated storage area of a storage device of the first server.

15. The method of claim 1, further comprising:

executing a monitoring process on the first server;

wherein monitoring the plurality of workloads comprises monitoring the first workload using the monitoring process.

16. The method of claim 15, further comprising storing server data regarding the plurality of servers in a data store coupled to a cloud control system, the server data including data received from the monitoring process, wherein the cloud control system is coupled to send messages to the first server, and wherein moving the first workload is performed at least in part based on the server data.

17. The method of claim 1, further comprising:

detecting, via the cloud control system, an end of the high demand placed by the first workload upon the first server, wherein the end of the high demand is within a predefined range of the average demand of the other workloads being handled by the first server; and

in response to detecting the end of the high demand, moving the first workload from the dedicated hardware resource to the first shared hardware resource.

18. A method for managing resources in a data center, the data center comprising a plurality of servers in a network, the plurality of servers comprising hardware resources for use by the data center, each respective server using a file system for file storage on the respective server, the method comprising:

providing, via the data center, a virtual machine for each of a plurality of users, each respective virtual machine to use a portion of the hardware resources of the data center, and wherein the hardware resources include storage resources distributed onto each of the plurality of servers;

monitoring, via an agent executing on each of the plurality of servers, a plurality of workloads being handled by each respective server, each workload being initiated based on a request by one of the virtual machines, each workload using one or more dedicated hardware resources, wherein each workload is allocated to the one or more hardware resources;

identifying, via an agent, a first workload requiring increased use of a storage resource of the hardware resources; and

provisioning the plurality of servers to temporarily allocate additional storage resources from the hardware resources to the first workload, the provisioning comprising moving the first workload of the plurality of servers from a shared storage resource to at least one dedicated storage resource of a first server of the plurality of servers, wherein moving the first workload comprises specifying to the file system of the first server that data storage for the first workload is to be handled by the dedicated storage resource,

wherein the at least one dedicated storage resource of the first server is only used by resources that are moved responsive to the identification of increased use.

19. The method of claim 18, wherein the first workload is being processed within a first respective virtual machine on the file system, and further comprising programming the file system of the first server from an application program interface (API) running on a computing device of one of the plurality of users.

20. The method of claim 18, wherein:

the shared storage resource is a first pool under the file system within which a plurality of virtual machines of other of the plurality of users are running; and

the dedicated storage resource is a second pool under the file system;

wherein the second pool stores at least a portion of files associated with the first workload prior to moving the first workload to the dedicated storage resource.

21. The method of claim 20, wherein the first pool comprises at least four drives in a RAID configuration, the second pool comprises at least two drives, and the second pool is operating as a mirror of the first pool.

22. The method of claim 18, wherein an agent on one of the plurality of servers identifies an increased demand of the first workload upon the shared storage resource.

23. The method of claim 18, wherein the file system comprises a combined file system and logical volume manager that implements snapshots.

24. A data center comprising:

a plurality of servers in a network;

each of the plurality of servers providing a virtual machine for each of a plurality of users, each respective virtual machine to use a portion of hardware resources of the data center, and wherein the hardware resources include storage and processing resources distributed onto each of the plurality of servers;

the plurality of servers communicating with one another via messages, at least a portion of the messages being sent from any respective server comprising status information regarding a hardware resource utilization status of that respective server;

each of the plurality of servers being configured to receive a request from a first of the respective virtual machines to handle a workload requiring increased use of the hardware resources, each workload associated with a virtual machine of one of the plurality of users and using one or more dedicated hardware resources, wherein each workload is allocated to the one or more hardware resources; and

the plurality of servers temporarily allocating additional resources from the hardware resources to the first respective virtual machine, the allocating based at least in part on status information provided by one or more of the messages, the allocating comprising moving the first workload of the plurality of servers from a shared storage resource to at least one dedicated storage resource of a first server of the plurality of servers, wherein moving the first workload comprises specifying to the file system of the first server that data storage for the first workload is to be handled by the dedicated storage resource,

wherein the at least one dedicated additional resource is only used by resources that are moved responsive to status information indicating increased use.