[PATCH 0/3]CFQ: add think time check for service tree and group

[PATCH 0/3]CFQ: add think time check for service tree and group

[Shaohua Li]

Hi,
  Currently when the last queue of a service tree/group is empty, CFQ
doesn't expire the queue. This is to allow requests from the tree/group
come soon, so tree/group doesn't miss its share. But if the think time
is big, the assumption isn't correct. idling the queue is just wasting
bandwidth.

Originally I was hoping this can resolve Vivek's fsync issue, but it
doesn't. The fsync issue is caused by queue idling. But since think time
check only helps for think time above default queue idle time (8ms),
think time check doesn't help.

On the other hand, think time check is still helpful for queues with
think time. I had test case in follow patches show throughput
improvement without sacrifice tree/group shares.

Thanks,
Shaohua

Re: [PATCH 0/3]CFQ: add think time check for service tree and group

[Shaohua Li]

2011/7/4 Shaohua Li <shaohua.li@intel.com>:
> Hi,
>  Currently when the last queue of a service tree/group is empty, CFQ
> doesn't expire the queue. This is to allow requests from the tree/group
> come soon, so tree/group doesn't miss its share. But if the think time
> is big, the assumption isn't correct. idling the queue is just wasting
> bandwidth.
>
> Originally I was hoping this can resolve Vivek's fsync issue, but it
> doesn't. The fsync issue is caused by queue idling. But since think time
> check only helps for think time above default queue idle time (8ms),
> think time check doesn't help.
>
> On the other hand, think time check is still helpful for queues with
> think time. I had test case in follow patches show throughput
> improvement without sacrifice tree/group shares.
>
Jens,
can you look at the patches? I refreshed some patches in replying
Vivek, I can resend if required.

Thanks,
Shaohua

Re: [PATCH 0/3]CFQ: add think time check for service tree and group

[Jens Axboe]

On 2011-07-11 07:44, Shaohua Li wrote:
> 2011/7/4 Shaohua Li <shaohua.li@intel.com>:
>> Hi,
>>  Currently when the last queue of a service tree/group is empty, CFQ
>> doesn't expire the queue. This is to allow requests from the tree/group
>> come soon, so tree/group doesn't miss its share. But if the think time
>> is big, the assumption isn't correct. idling the queue is just wasting
>> bandwidth.
>>
>> Originally I was hoping this can resolve Vivek's fsync issue, but it
>> doesn't. The fsync issue is caused by queue idling. But since think time
>> check only helps for think time above default queue idle time (8ms),
>> think time check doesn't help.
>>
>> On the other hand, think time check is still helpful for queues with
>> think time. I had test case in follow patches show throughput
>> improvement without sacrifice tree/group shares.
>>
> Jens,
> can you look at the patches? I refreshed some patches in replying
> Vivek, I can resend if required.

Please resend the series after a refresh like that, otherwise I tend to
lose track of the various things flying around.

-- 
Jens Axboe

Re: [PATCH 0/3]CFQ: add think time check for service tree and group

[Vivek Goyal]

On Mon, Jul 11, 2011 at 01:44:05PM +0800, Shaohua Li wrote:
> 2011/7/4 Shaohua Li <shaohua.li@intel.com>:
> > Hi,
> >  Currently when the last queue of a service tree/group is empty, CFQ
> > doesn't expire the queue. This is to allow requests from the tree/group
> > come soon, so tree/group doesn't miss its share. But if the think time
> > is big, the assumption isn't correct. idling the queue is just wasting
> > bandwidth.
> >
> > Originally I was hoping this can resolve Vivek's fsync issue, but it
> > doesn't. The fsync issue is caused by queue idling. But since think time
> > check only helps for think time above default queue idle time (8ms),
> > think time check doesn't help.
> >
> > On the other hand, think time check is still helpful for queues with
> > think time. I had test case in follow patches show throughput
> > improvement without sacrifice tree/group shares.
> >
> Jens,
> can you look at the patches? I refreshed some patches in replying
> Vivek, I can resend if required.

Yep, please resend. That would be easier.

Thanks
Vivek