VM Migration Using Page State Bitmaps for Externalized Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual machine (VM) migration techniques face inefficiencies due to unnecessary page copy operations, increased network traffic, and memory pressure issues, particularly when transferring externalized memory pages, which can lead to high latency and downtime during the migration process.
Innovation Solution
The method involves generating a page transfer state for local and externalized memory pages, allowing for the transfer of unique identifiers instead of page content, and using a page state bitmap to manage page states, thereby reducing unnecessary operations and optimizing memory access during VM migration across compute nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all memory pages (including externalized pages) are transferred during VM migration, then the migrated VM can access all required memory pages, but network traffic and CPU load increase significantly
Solution Approach 1:
The patent extracts only the necessary information (page identifiers and state bits) from the memory pages during migration, rather than transferring the entire page contents. The page state bitmap stores metadata about each page's location and transfer status, allowing the system to reconstruct memory state without moving all actual data, thus reducing network traffic while maintaining memory access completeness.
Solution Approach 2:
The patent introduces a page state bitmap as an intermediary data structure that tracks the state of memory pages during migration. This bitmap acts as a mediator between the source and destination VMs, recording which pages have been transferred, which are externalized, and which remain local, enabling efficient memory management without requiring complete page transfers.
2Speed
If externalized memory pages are transferred back to the source node, then the source node can access them locally, but memory pressure on the source node increases
Solution Approach 1:
The patent applies local quality by allowing different memory pages to have different locations and access characteristics. The page state bitmap tracks which pages are local to the source node, which are externalized to other nodes, and which are in transit. This enables the system to optimize access speed for local pages while maintaining the ability to access externalized pages through the distributed memory architecture, without forcing all pages to reside locally.
3Reliability
If VM migration is performed with complete page transfer, then the migrated VM has full memory access, but downtime during migration increases
Solution Approach 1:
The patent performs preliminary actions by pre-transferring memory pages and updating the page state bitmap before the actual VM migration occurs. The system proactively moves pages and tracks their state in advance, so that when the VM is migrated, the destination node already has the necessary page information available, reducing the downtime required for memory setup after migration.
Solution Approach 2:
The patent maintains continuity of useful action by allowing the VM to continue running on the source node while memory pages are being transferred and the page state bitmap is being updated. The migration process occurs concurrently with VM operation, and the destination node begins receiving page information before the VM stops, ensuring that memory access setup continues uninterrupted throughout the migration process.
Data Source
AI summary
A method includes running a Virtual Machine (VM) on a first compute node in a plurality of compute nodes that communicate with one another over a communication network. The VM is migrated from the first compute node to a second compute node in the plurality by generating, for memory pages accessed by the VM, page transfer state of one or more local memory pages that are accessed locally on the first compute node, and of one or more externalized memory pages whose access is not confined to the first node. Based on the page transfer state, the migrated VM is provided with access to the memory pages, including both the local and the externalized memory pages, on the second compute node.