Hi, Paolo,
A few months ago, Juan told me that this series requires your or someone
familiar with memory API's feedback.
Could you please review it and provide some suggestions?
On 2023/3/17 下午4:18, Chuang Xu wrote:
> In this version:
>
> - delete useless line change.
> - update comments and commit messages.
>
> The duration of loading non-iterable vmstate accounts for a significant
> portion of downtime (starting with the timestamp of source qemu stop and
> ending with the timestamp of target qemu start). Most of the time is spent
> committing memory region changes repeatedly.
>
> This patch packs all the changes to memory region during the period of
> loading non-iterable vmstate in a single memory transaction. With the
> increase of devices, this patch will greatly improve the performance.
>
> Here are the test1 results:
> test info:
> - Host
> - Intel(R) Xeon(R) Platinum 8362 CPU
> - Mellanox Technologies MT28841
> - VM
> - 32 CPUs 128GB RAM VM
> - 8 16-queue vhost-net device
> - 16 4-queue vhost-user-blk device.
>
> time of loading non-iterable vmstate downtime
> before 112 ms 285 ms
> after 20 ms 194 ms
>
>
> In test2, we keep the number of the device the same as test1, reduce the
> number of queues per device:
>
> Here are the test2 results:
> test info:
> - Host
> - Intel(R) Xeon(R) Platinum 8362 CPU
> - Mellanox Technologies MT28841
> - VM
> - 32 CPUs 128GB RAM VM
> - 8 1-queue vhost-net device
> - 16 1-queue vhost-user-blk device.
>
> time of loading non-iterable vmstate downtime
> before 65 ms 151 ms
>
> after 19 ms 100 ms
>
>
> In test3, we keep the number of queues per device the same as test1, reduce
> the number of devices:
>
> Here are the test3 results:
> test info:
> - Host
> - Intel(R) Xeon(R) Platinum 8362 CPU
> - Mellanox Technologies MT28841
> - VM
> - 32 CPUs 128GB RAM VM
> - 1 16-queue vhost-net device
> - 1 4-queue vhost-user-blk device.
>
> time of loading non-iterable vmstate downtime
> before 24 ms 51 ms
> after 9 ms 36 ms
>
>
> As we can see from the test results above, both the number of queues and
> the number of devices have a great impact on the time of loading non-iterable
> vmstate. The growth of the number of devices and queues will lead to more
> mr commits, and the time consumption caused by the flatview reconstruction
> will also increase.
>
> Please review, Chuang
>
Thanks!