Migrating processes between source host and destination host using a shared virtual file system

Patent metadata

Field	Value
Publication number	US-9529616-B2
Application number	US-201314101498-A
Country	US
Kind code	B2
Filing date	Dec 10, 2013
Priority date	Dec 10, 2013
Publication date	Dec 27, 2016
Grant date	Dec 27, 2016

Abstract

A process can be scheduled between first and second hosts that using a virtual file system that is shared between the hosts can be used. The process, running on a first hypervisor of the first host, can be scheduled to run on a second hypervisor of the second host. A file can be created that includes the data content of the process address space for the file. The file can be mapped address space of the virtual file system. Data from the physical memory of the first host can be transferred to physical memory of the second host using page fault routines.

First claim

What is claimed is:

1. A method for scheduling a process between first and second hosts that use a virtual file system, the method comprising:
scheduling the process, running on a first hypervisor of the first host, to run on a second hypervisor of the second host, wherein the first hypervisor includes a first page table, a first list of file locations, and a first distributed shared memory (DSM) module, and wherein the second hypervisor includes a second page table, a second list of file locations, and a second DSM module;
identifying a process address space for the process, wherein the process address space holds data content;
creating a file that maps to the process address space;
mapping the file to address space of the virtual file system that is shared between the first host and the second host using an mmap( )system call;
running the process on the second hypervisor;
generating a page fault, in response to the process requesting access to the file in the virtual file system and by referencing the second page table, wherein the second page table indicates that data content of the file is not stored in physical memory of the second host;
determining, by the second DSM module, that the file access request is for data that is stored in physical memory of the first host by referencing the second list in the second hypervisor, wherein referencing the second list does not modify the second page table; and
transferring, by the first DSM module in connection with the second DSM module, in response to the determining, the data from the physical memory of the first host to the physical memory of the second host, wherein the second page table is updated with page associated with transferred data, and wherein the first and second page tables are not modified to include additional information or bits to indicate whether or not the page is stored at a remote host.

2. The method of claim 1 , further comprising generating the second list on the second hypervisor, that specifies whether pages corresponding to memory of the process address space are stored on the second host and wherein determining that the file access request is for data that is stored in the physical memory of the first host is based upon the second list.

3. The method of claim 1 , wherein transferring the data from the physical memory of the first host to the physical memory of the second host includes transferring the data using a network port connection between the first hypervisor and the second hypervisor.

4. The method of claim 2 , further comprising creating the first list, on the first hypervisor, that specifies whether pages corresponding to memory of the process address space are stored on the first host, wherein creating and modifying the first list does not include modifying the first page table.

5. The method of claim 4 , further comprising updating the first list by removing a page entry corresponding to the data transferred from the physical memory of the first host.

6. The method of claim 1 , wherein the first and second hypervisors are respective Linux kernels.

7. The method of claim 6 , wherein the virtual file system corresponds to another Linux kernel running in parallel on each of the first and second hypervisors.

Assignees

• IBM

Inventors

• Borntraeger Christian
• Carstens Heiko
• Dingel Dominik
• Klein Matthias
• Lueck Einar

Classifications

• G06F2009/45595

  Network integration; Enabling network access in virtual machine instances · CPC title

• H04L67/1008

  based on parameters of servers, e.g. available memory or workload (monitoring of computer activity G06F11/30) · CPC title

• G06F2009/45583

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

• G06F9/45558 Primary

  Hypervisor-specific management and integration aspects · CPC title

• G06F9/4856 Primary

  resumption being on a different machine, e.g. task migration, virtual machine migration (G06F9/5088 takes precedence) · CPC title