Method for targeted resource virtualization in containers

Patent metadata

Field	Value
Publication number	US-9348622-B2
Application number	US-201414251989-A
Country	US
Kind code	B2
Filing date	Apr 14, 2014
Priority date	Nov 26, 2013
Publication date	May 24, 2016
Grant date	May 24, 2016

Abstract

A method and computer program product for targeted container virtualization, where only separate components of a computer system or a server are virtualized. The OS kernel and other server resources are not virtualized. Only selected components—applications or resources are targeted for virtualization instead of virtualization of the entire system. Targeted virtualization provides for more flexible container isolation from each other and from a host node. This, in turn, provides for optimized more flexible cloud infrastructure. Each element within a container virtualization model is optional in terms of virtualization. The element's virtualization option can be turned on and off by an administrator or by a client who owns the container.

First claim

What is claimed is:

1. A system for targeted virtualization in containers, the system comprising:
a host hardware node having a host OS kernel;
a plurality of host OS kernel objects implemented on the host hardware node; and
at least one container running on the host hardware node, the container virtualizing the host OS and using selected host OS kernel utilities,
wherein:
the host OS kernel utilities have a virtualization on-off switch;
the selected host OS kernel object is virtualized inside the container if the utility virtualization switch is on; and
the container uses the host OS kernel objects shared with other containers running on the hardware node.

2. The system of claim 1 , wherein the utility virtualization switch is turned on by a container administrator.

3. The system of claim 1 , wherein the utility virtualization switch is turned on based on a container user requirements.

4. The system of claim 1 , wherein the host OS kernel objects are not virtualized.

5. The system of claim 1 , wherein the host OS kernel objects are shared among containers running on the hardware node.

6. The system of claim 1 , wherein the container uses selected virtualized host OS kernel objects and shared host OS kernel objects based on user requirements.

7. The system of claim 1 , wherein the host OS kernel is patched for selected virtualization of utilities in a form of separate utility modules.

8. The system of claim 1 , wherein the host OS kernel objects used for selected virtualization are any of:
memory;
I/O operations;
disk;
network;
users and groups;
devices;
PID tree;
IPC objects;
user applications; and
system modules.

9. The system of claim 1 , wherein the container is a part of a cloud-based infrastructure.

10. A computer-implemented method for targeted virtualization in a container, the method comprising:
(a) launching an OS kernel on a host hardware node;
(b) patching the host OS kernel for selected virtualization of host OS kernel objects;
(c) activating a utility virtualization on-off switch on the host OS kernel objects;
(d) selecting the host OS kernel objects to be virtualized by turning the utility virtualization switch on;
(e) launching a first container on the hardware node, the first container virtualizing the host OS;
(f) virtualizing the selected host OS kernel objects inside the container;
(g) repeating steps (b)-(f) for another container launched on the hardware node, wherein the first container shares the host OS kernel objects that are not virtualized with other containers running on the hardware node.

11. The method of claim 10 , wherein the selecting of the host OS kernel objects is implemented based on user requirements.

12. The method of claim 10 , wherein the containers cannot access objects virtualized inside the other containers.

13. The method of claim 10 , wherein the containers form a cloud infrastructure.

14. The method of claim 10 , further comprising creating a beancounter for controlling resource usage by container processes.

15. The method of claim 14 , wherein virtualized container resources are controlled by the beancounter.

16. A computer-implemented method for backup optimization in a dedicated container, the method comprising:
(a) launching an OS kernel on a host hardware node;
(b) patching the host OS kernel for selected virtualization of host OS kernel objects;
(c) activating a utility virtualization on-off switch on the host OS kernel objects;
(d) selecting the host OS kernel objects to be virtualized by turning the utility virtualization switch on;
(e) launching a dedicated backup container on the hardware node, the backup container virtualizing the host OS;
(f) virtualizing the selected host OS kernel objects inside the dedicated container; and
(g) starting a host backup utility configured to backup container data,
wherein the host backup utility only backs up the host OS kernel objects that are not virtualized within the dedicated backup container.

17. A system for targeted virtualization in a container, the system comprising:
a processor;
a memory coupled to the processor; and
a computer program logic stored in the memory and executed on the processor, the computer program logic for implementing the steps (a)-(g) of claim 10 .

Assignees

• Emelyanov Pavel
• Petrov Igor
• Protassov Stanislav S
• Beloussov Serguei M
• Parallels IP Holdings GmbH

Inventors

• Emelyanov Pavel
• Petrov Igor
• Protassov Stanislav S
• Beloussov Serguei M

Classifications

• G06F9/4555

  Para-virtualisation, i.e. guest operating system has to be modified · CPC title

• G06F9/455 Primary

  Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines · CPC title