On Tue, Aug 08, 2017 at 04:53:08PM +0200, Alberto Garcia wrote:
>On Tue 08 Aug 2017 03:45:44 PM CEST, Manos Pitsidianakis wrote:
>> On Tue, Aug 08, 2017 at 03:13:36PM +0200, Alberto Garcia wrote:
>>>On Mon 31 Jul 2017 11:54:41 AM CEST, Manos Pitsidianakis wrote:
>>>> block/throttle.c uses existing I/O throttle infrastructure inside a
>>>> block filter driver. I/O operations are intercepted in the filter's
>>>> read/write coroutines, and referred to block/throttle-groups.c
>>>>
>>>> The driver can be used with the syntax
>>>> -drive driver=throttle,file.filename=foo.qcow2, \
>>>>         limits.iops-total=...,throttle-group=bar
>>>
>>>Sorry for not having noticed this earlier, but can't you define the
>>>throttling group (and its limits) using -object throttle-group ... as
>>>shown in the previous patch, and simply reference it here? Or would we
>>>have two alternative ways of setting the throttling limits?
>>>
>>>What happens if you have many -drive lines each one with a different set
>>>of limits but with the same throttling group?
>>
>> The limits of the last one to be processed will win.
>
>That's what the current throttling API does, and I tend to agree that
>it's not completely straightforward (a few people have asked me the same
>question since this feature is available).
>
>If we're going to add a new API we could eliminate this ambiguity. After
>all the previous -drive throttling.iops-total=... would still be
>available, wouldn't it?

Indeed, it already is.

>
>> So basically if we have anonymous groups, we accept limits in the
>> driver options but only without a group-name.
>
>In the commit message you do however have limits and a group name, is
>that a mistake?
>
>    -drive driver=throttle,file.filename=foo.qcow2, \
>           limits.iops-total=...,throttle-group=bar
>

Sorry this wasn't clear, I'm actually proposing to remove limits from 
the throttle driver options and only create/config throttle groups via 
-object/object-add.