Re: [PATCH v2] migration/throttle: Add cpu-throttle-tailslow migration parameter

[Eric Blake]

On 3/15/20 11:29 PM, Keqian Zhu wrote:
> At the tail stage of throttling, the Guest is very sensitive to
> CPU percentage while the @cpu-throttle-increment is excessive
> usually at tail stage.
>
> If this parameter is true, we will compute the ideal CPU percentage
> used by the Guest, which may exactly makes dirty rate to be dirty
> rate threshold. Then we will choose a smaller throttle increment
> between the one specified by @cpu-throttle-increment and the one
> generated by ideal CPU percentage.
>
> Therefore, it is compatible to traditional throttling, meanwhile
> the throttle increment won't be excessive at tail stage. This may
> make migration time longer, and is disabled by default.
>
> Signed-off-by: Keqian Zhu <address@hidden>
> ---
> Cc: Juan Quintela <address@hidden>
> Cc: "Dr. David Alan Gilbert" <address@hidden>
> Cc: Eric Blake <address@hidden>
> Cc: Markus Armbruster <address@hidden>
> ---

> +++ b/qapi/migration.json
> @@ -552,6 +552,21 @@
>   #                          auto-converge detects that migration is not making
>   #                          progress. The default value is 10. (Since 2.7)
>   #
> +# @cpu-throttle-tailslow: Make CPU throttling slower at tail stage
> +#                         At the tail stage of throttling, the Guest is very
> +#                         sensitive to CPU percentage while the @cpu-throttle
> +#                         -increment is excessive usually at tail stage.
> +#                         If this parameter is true, we will compute the ideal
> +#                         CPU percentage used by the Guest, which may exactly
> +#                         makes dirty rate to be dirty rate threshold. Then we

Grammar is off here, but I don't know if you meant "which may exactly 
make the dirty rate match the dirty rate threshold" or something else.

> +#                         will choose a smaller throttle increment between the
> +#                         one specified by @cpu-throttle-increment and the one
> +#                         generated by ideal CPU percentage.
> +#                         Therefore, it is compatible to traditional 
> throttling,
> +#                         meanwhile the throttle increment won't be excessive
> +#                         at tail stage.
> +#                         The default value is false. (Since 5.0)

Is this a bug-fix that must make it into 5.0?  It seems more like a 
feature addition, at which point listing it for 5.1 is probably a better 
idea.

--
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org