Memory migration method, apparatus, and computing device
Embodiments of this application provide a memory migration method, an apparatus, and a computing device. The method includes: creating a processing thread, a sending thread, and a traversal thread for a memory region; where the traversal thread iteratively executes a traversal operation, where the traversal operation includes traversing a dirty bitmap of the memory region and assigning a dirty page address to a corresponding processing thread; the processing thread iteratively executes a processing operation, where the processing operation includes reading dirty page data based on the assigned dirty page address, and adding round number flag information to the dirty page data to generate a to-be-migrated dirty page; and the sending thread iteratively executes a transmission operation, where the transmission operation includes sending the to-be-migrated dirty page to a destination end, and the destination end restores respective dirty page data to destination memories based on iteration round numbers in different to-be-migrated dirty pages. The embodiments of this application improve memory migration efficiency.
1. A memory migration method, comprising:
creating a processing thread, a sending thread, and a traversal thread for a memory region; wherein:
the traversal thread iteratively executes a traversal operation, wherein the traversal operation comprises traversing a dirty bitmap of the memory region and assigning a dirty page address to a corresponding processing thread,
the processing thread iteratively executes a processing operation, wherein the processing operation comprises reading dirty page data based on the assigned dirty page address, and adding round number flag information to the dirty page data to generate a to-be-migrated dirty page, and
the sending thread iteratively executes a transmission operation, wherein the transmission operation comprises sending the to-be-migrated dirty page to a destination end.
2. The method according to claim 1 , wherein the processing thread specifically iteratively executes the following processing operation:
reading dirty page data based on the assigned dirty page address;
compressing the dirty page data; and
adding round number flag information for identifying a corresponding iteration round number to the compressed dirty page data to generate a to-be-migrated dirty page.
3. The method according to claim 1 , further comprising:
creating a synchronization thread for the memory region; wherein
the synchronization thread iteratively executes a synchronization operation, wherein the synchronization operation comprises acquiring a dirty bitmap of the memory region and sending the dirty bitmap to the traversal thread.
4. The method according to claim 1 , wherein a plurality of processing threads are specifically created, and the plurality of processing threads run in parallel and iteratively execute the processing operation separately;
the traversal thread specifically iteratively executes the following traversal operation:
traversing a dirty bitmap of the memory region to determine a dirty page address and an iteration round number;
assigning, based on the number of to-be-processed dirty pages, corresponding to-be-processed tasks to one or more processing threads with free resources, wherein the to-be-processed task comprises the iteration round number and at least one to-be-processed dirty page address.
5. The method according to claim 1 , wherein the processing thread specifically iteratively executes the following processing operation:
reading dirty page data based on the assigned dirty page address;
adding round number flag information for identifying an iteration round number to the dirty page data to generate a to-be-migrated dirty page; and
caching the to-be-migrated dirty page to a dirty page cache page corresponding to the processing thread.
6. The method according to claim 5 , wherein the sending thread specifically iteratively executes the following sending operation:
transmitting a to-be-migrated dirty page in the dirty page cache queue to the destination end.
7. The method according to claim 1 , wherein a source memory is divided into a plurality of the memory regions, and different memory regions do not overlap each other; and
wherein the creating the processing thread, the sending thread, and the traversal thread for the memory region comprises:
creating a traversal thread for the memory region, at least one sending thread, and at least one processing thread.
8. The method according to claim 1 , wherein the creating the sending thread, the processing thread, and the traversal thread for the memory region comprises:
creating a sending thread, a processing thread, and a traversal thread for a memory region in one or more processing components.
9. The method according to claim 1 , wherein a plurality of processing threads and a plurality of sending threads are specifically created; and
wherein the plurality of processing threads run in parallel, and iteratively execute the processing operation separately, and the plurality of sending threads run in parallel, and iteratively execute the transmission operation separately.
10. A memory migration method, comprising:
creating a sending thread and a traversal thread for a memory region; wherein
the traversal thread iteratively executes a traversal operation, wherein the traversal operation comprises traversing a dirty bitmap of the memory region, reading dirty page data based on a dirty page address, and adding round number flag information to the dirty page data to generate a to-be-migrated dirty page, and
the sending thread iteratively executes a transmission operation, wherein the transmission operation comprises sending the to-be-migrated dirty page to a destination end.
11. The method according to claim 10 , further comprising:
creating a synchronization thread for the memory region; wherein
the synchronization thread iteratively executes a synchronization operation, wherein the synchronization operation comprises acquiring a dirty bitmap of the memory region and sending the dirty bitmap and the iteration round number to the traversal thread.
12. The method according to claim 10 , wherein a plurality of sending threads are specifically created, and the plurality of sending threads run in parallel and iteratively execute the sending operation separately; and
the traversal thread specifically iteratively executes the following traversal operation:
traversing a dirty bitmap of the memory region to determine a dirty page address and an iteration round number;
reading dirty page data based on the dirty page address;
adding round number flag information for identifying the iteration round number to the dirty page data to generate a to-be-migrated dirty page; and
separately assigning, based on the number of to-be-processed dirty pages, at least one to-be-migrated dirty page to one or more sending threads with free resources.
13. The method according to claim 10 , wherein a source memory is divided into a plurality of the memory regions, and different memory regions do not overlap each other; and
the creating the sending thread and the traversal thread for the memory region comprises:
creating for the memory region a traversal thread and a sending thread.
14. The method according to claim 10 , wherein the creating the sending thread and the traversal thread for the memory region comprises:
creating a sending thread and a traversal thread for a memory region in one or more processing components.
15. The method according to claim 10 , wherein a plurality of the sending threads are specifically created; and
a plurality of sending threads run in parallel, and iteratively execute the transmission operation separately.