Distributed systems having versioned storage replica tables

Patent metadata

Field	Value
Publication number	US-2025036441-A1
Application number	US-202318358155-A
Country	US
Kind code	A1
Filing date	Jul 25, 2023
Priority date	Jul 25, 2023
Publication date	Jan 30, 2025
Grant date	—

Abstract

A process includes creating, by a kernel of a distributed system, entries in a storage replica table, which are associated with a guest physical memory address. The process includes storing, by the kernel, data in the entries associating the entries with respective versions of content for the guest physical memory address; storing, by the kernel, data in the entries associating the entries with respective stable storage block addresses; and storing, by the kernel, data in the entries associating the entries with respective real physical memory addresses. The process includes, responsive to a read request to read content associated with a first version of the versions, accessing, by the kernel, the storage replica table. The process includes, responsive to the accessing the storage replica table, identifying, by the kernel, a first entry of the entries associated with the first version. The first entry contains data associating the first entry with a first real physical memory address. The process includes reading, by the kernel, data from the first real physical memory address.

First claim

What is claimed is:

1 . A distributed system comprising:
a plurality of computer nodes; wherein a first computer node of the plurality of computer nodes comprises:
a program that provides a first request to write a first content to a guest virtual memory address and a second request to write a second content to the guest virtual memory address;
a guest operating system that:
responsive to the first request, provides a third request to write the first content associated with a first version to a guest physical memory address; and
responsive to the second request, provides a fourth request to write the second content associated with a second version to the guest physical memory address; and

a kernel that:
responsive to the third request, stores a first entry in a storage replica table associating the guest physical memory address with a first real physical memory address, a first block address of stable storage and the first version; and
responsive to the fourth request, stores a second entry in the storage replica table associating the guest physical memory address with a second real physical memory address, a second block address of stable storage and the second version.

2 . The distributed system of claim 1 , wherein the kernel, responsive to the third request, writes the first content to the first real physical memory address and the first block address of stable storage.

3 . The distributed system of claim 2 , wherein the kernel, responsive to the second request, modifies the second entry to designate the second entry as being immutable.

4 . The distributed system of claim 1 , wherein the kernel, responsive to the fourth request, writes the second content to the second real physical memory address and the second block address of stable storage.

5 . The distributed system of claim 1 , wherein the kernel, responsive to a boot-up of a first computer node of the plurality of computer nodes and responsive to a request to load a block from the stable storage during the boot-up:
accesses the storage replica table; reads the second entry from the storage replica table; determines, based on the second entry, that content stored at the second real physical memory address is a clone of the block; and handles the request to load the block responsive to the determination.

6 . The distributed system of claim 1 , wherein the kernel regulates how often entries corresponding to newer versions of guest memory pages may be added to the storage replica table based on a snapshot generation policy.

7 . The distributed system of claim 1 , further comprising:
a plurality of hyper-kernels distributed over the plurality of computer nodes and forming the kernel; and a virtual machine distributed over the plurality of computer nodes and containing the guest operating system.

8 . The distributed system of claim 1 , further comprising:
a hypervisor hosted by the first computer node and forming the kernel; and a virtual machine hosted by the first computer node and containing the guest operating system.

9 . A method comprising:
creating, by a kernel of a distributed system, entries in a storage replica table associated with a guest physical memory address; storing, by the kernel, data in the entries associating the entries with respective versions of content for the guest physical memory address; storing, by the kernel, data in the entries associating the entries with respective stable storage block addresses; storing, by the kernel, data in the entries associating the entries with respective real physical memory addresses; and responsive to a read request to read content associated with a first version of the versions:
accessing, by the kernel, the storage replica table;
responsive to accessing the storage replica table, identifying, by the kernel, a first entry of the entries associated with the first version, the first entry containing data associating the first entry with a first real physical memory address of the plurality of real physical memory addresses; and
reading, by the kernel, data from the first real physical memory address.

10 . The method of claim 9 , further comprising creating, by the kernel, additional entries in the storage replica table associated with additional guest physical memory addresses.

11 . The method of claim 9 , further comprising:
storing, by the kernel, data in the first entry designating the first entry as being immutable.

12 . The method of claim 9 , further comprising:
providing, by a program instance, the read request; and restoring, by the program instance, content for a guest virtual memory corresponding to the first version.

13 . The method of claim 9 , further comprising:
providing, by a program instance, the read request; and comparing, by the program instance, first content for a guest virtual memory corresponding to the first version to second content for the guest virtual memory corresponding to a second version of the versions.

14 . The method of claim 9 , further comprising:
providing, by a program instance, the read request; determining, by the program instance, changes between first content for a guest virtual memory address corresponding to the first version and second content for the guest virtual memory address corresponding to a second version of the versions; correlating, by the program instance, the changes with a system event; and detecting, by the program instance, a security attack based on the correlation.

15 . The method of claim 9 , further comprising:
creating a snapshot of the guest physical memory address associated with the first version, wherein the creating comprise mirroring the data read from the first real physical memory address to a second real physical memory address.

16 . A non-transitory machine-readable storage medium that stores instructions that, when executed by a first compute node of a distributed system, cause the first computer node to, responsive to a request to write content to a guest physical memory address:
access a storage replica table, wherein the storage replica table comprises a first entry corresponding to a first snapshot for the guest physical memory address; create a second entry for the storage replica table; store data in the second entry designating the second entry as corresponding to a second snapshot for the guest physical memory address; store data in the second entry associating the guest physical memory address with a block address of stable storage; store data in the second entry associating the guest physical memory address with a real physical memory address; and write the content to the real physical memory address.

17 . The storage medium of claim 16 , wherein the instructions, when executed by the first compute node further cause the first computer node to, responsive to a boot-up of the first computer node:
responsive to a request to load a block from the stable storage to a second guest physical memory address associated with the first computer node, access the storage replica table to determine whether a clone of the block is stored at a second real physical memory address; and process the request to load the block from the stable storage based on a result of the determination of whether the clone of the block is stored at the second real physical memory address.

18 . The storage medium of claim 16 , wherein the instructions, when executed by the first compute node, further cause the first computer node to, responsive to the request, write the content to the bloc

Assignees

• Hewlett Packard Entpr Dev Lp

Inventors

• Berman Michael
• Reynolds Keith

Classifications

• G06F9/45558 Primary

  Hypervisor-specific management and integration aspects · CPC title

• G06F2009/45595

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

• G06F2009/45583

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

• G06F9/45545 Primary

  Guest-host, i.e. hypervisor is an application program itself, e.g. VirtualBox · CPC title