Guest ordering of host file system writes

Patent metadata

Field	Value
Publication number	US-9836402-B1
Application number	US-201615216826-A
Country	US
Kind code	B1
Filing date	Jul 22, 2016
Priority date	Jul 22, 2016
Publication date	Dec 5, 2017
Grant date	Dec 5, 2017

Abstract

Systems and methods for data storage management technology that enables a guest module of a virtual machine to indicate an order in which a host module should write data from physical memory to a secondary storage. An example method may comprise: identifying, by a processing device executing a host module, a plurality of modifications to physical memory made by a plurality of direct access operations executed by a guest module of a virtual machine; determining, by the host module, an order of the plurality of modifications to physical memory; receiving, by the host module, a synchronization request from the guest module; and responsive to the synchronization request, copying, by the host module, data from the physical memory to a secondary storage in view of the order of the plurality of modifications.

First claim

What is claimed is:

1. A method comprising:
identifying, by a processing device executing a host kernel, a plurality of modifications to physical memory made by a plurality of direct access operations, wherein the direct access operations are executed by a guest operating system and bypass the host kernel;
receiving, by the host kernel, information from the guest operating system indicating an order of the plurality of modifications made to the physical memory;
receiving, by the host kernel, a synchronization request from the guest operating system; and
responsive to the synchronization request, copying, by the host kernel, data from the physical memory to a secondary storage in view of the order of the plurality of modifications.

2. The method of claim 1 , wherein the guest operating system comprises a guest kernel executing on a virtual machine and the host kernel comprises a hypervisor managing the virtual machine.

3. The method of claim 2 , wherein the plurality of direct access operations enable the guest operating system to bypass the hypervisor and execute a load instruction of the physical memory.

4. The method of claim 1 , wherein the physical memory comprises non-volatile direct access memory, and wherein a user space process running on the guest operating system is to bypass a kernel of the guest operating system and to modify the non-volatile direct access memory.

5. The method of claim 1 , wherein the host kernel is to refrain from performing the copying of the physical memory to the secondary storage until receiving the synchronization request from the guest operating system.

6. The method of claim 1 , wherein the host kernel supports multiple guest operating systems that perform modifications to a file system residing in the physical memory, wherein the modifications are stored in a shared page cache provided by the host kernel without being stored in a page cache of any of the guest operating systems.

7. The method of claim 1 , further comprising:
monitoring, by the host kernel, a quantity of unsynchronized physical memory, wherein the unsynchronized physical memory comprises data that is modified by the guest operating system and is not synchronized with the secondary storage; and
permitting the guest operating system to continue modifying the physical memory when the quantity is within a threshold quantity.

8. The method of claim 7 , further comprising: in response to the quantity of unsynchronized physical memory exceeding a threshold, notifying the guest operating system that additional modifications to the physical memory are restricted.

9. The method of claim 7 , wherein the threshold quantity is in view of a quantity of dirty memory pages associated with a particular guest operating system or all guest operating systems hosted by the host kernel.

10. The method of claim 1 , wherein the order of the plurality of modifications indicates an order with respect to the synchronization request and identifies one or more of the plurality of modifications that occurred prior to the synchronization request being issued.

11. The method of claim 10 , wherein the order of the plurality of modifications further indicates an order between one or more of the plurality of modifications that occurred prior to the synchronization request being issued.

12. The method of claim 1 , wherein determining the order of the plurality of modifications, comprises receiving, from the guest operating system, a data structure comprising information indicating the order.

13. The method of claim 12 , wherein the data structure specifies a memory barrier that permits a first modification that occurred before the memory barrier to be synchronized to secondary storage and limits a second modification that occurred after the memory barrier from being synchronized to the secondary storage.

14. The method of claim 1 , wherein the order of the plurality of modifications to physical memory is used to generate a snapshot of the memory occupied by the guest operating system.

15. A system comprising:
a memory; and
a processing device operatively coupled to the memory, the processing device to:
identify, using a host kernel, a plurality of modifications to physical memory made by a plurality of direct access operations, wherein the direct access operations are executed by a guest operating system and bypass a kernel of the host kernel;
receive, using the host kernel, information from the guest operating system indicating an order of the plurality of modifications made to the physical memory;
receive, using the host kernel, a synchronization request from the guest operating system; and
responsive to the synchronization request, copy, by the host kernel, data from the physical memory to a secondary storage in view of the order of the plurality of modifications.

16. The system of claim 15 , wherein the guest operating system comprises a guest kernel executing on a virtual machine and the host kernel comprises a hypervisor managing the virtual machine.

17. The system of claim 16 , wherein the plurality of direct access operations enable the guest operating system to bypass the hypervisor and execute a load instruction of the physical memory.

18. A non-transitory machine-readable storage medium storing instructions that cause a processing device to:
initiate, by a guest operating system of a virtual machine, a plurality of modifications to physical memory using a plurality of direct access operations, wherein the direct access operations bypass a host kernel;
provide, by the guest operating system, information to the host kernel indicating an order of the plurality of modifications; and
transmit, by the guest operating system, a synchronization request to initiate the host kernel to copy data from the physical memory to a secondary storage in view of the order of the plurality of modifications.

19. The non-transitory machine-readable storage medium of claim 18 , wherein the guest operating system synchronously initiates the synchronization request and waits for the host kernel to copy the data from the physical memory to the secondary storage.

20. The non-transitory machine-readable storage medium of claim 18 , wherein the guest operating system asynchronously initiates the synchronization request without waiting for the host kernel to copy the data from the physical memory to the secondary storage.

Assignees

• Red Hat Inc

Inventors

• Riel Henri Van

Classifications

• G06F2212/311

  In host system · CPC title

• G06F2212/152

  Virtualized environment, e.g. logically partitioned system · CPC title

• G06F12/0804

  with main memory updating (G06F12/0806 takes precedence) · CPC title

• G06F2212/1032

  Reliability improvement, data loss prevention, degraded operation etc · CPC title

• G06F2009/45583

  Memory management, e.g. access or allocation · CPC title