Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
about definition of iowait
8 views
Skip to first unread message
Cao jin
Jun 7, 2016, 7:20:07 AM6/7/16
to
Hi,
I noticed some problems about iowait entry of /proc/stat: it seems not
accurate, and sometimes will decrease in SMP.
For UP, we have a simple definition about iowait, which is:
The time that the processor is idle, during which there is a task
waiting for I/O.
This definition seems don't fit well when step into SMP era, so I think
maybe it can be extended like:
For a given CPU, the I/O wait time is the time during which that CPU was
idle, and there is at least one outstanding disk I/O operation requested
by a task scheduled on that CPU.
But tasks can migrate among cpus, so this definition maybe not accurate.
I also saw some suggestions that it should be a global value, not per
cpu, but don`t see clear definition, so anyone have suggestions about
it? (personally guessing, maybe: there is task waiting for outstanding
I/O while all cpus are idle? But I don't think it is good...)
--
Yours Sincerely,
Cao jin
I noticed some problems about iowait entry of /proc/stat: it seems not
accurate, and sometimes will decrease in SMP.
For UP, we have a simple definition about iowait, which is:
The time that the processor is idle, during which there is a task
waiting for I/O.
This definition seems don't fit well when step into SMP era, so I think
maybe it can be extended like:
For a given CPU, the I/O wait time is the time during which that CPU was
idle, and there is at least one outstanding disk I/O operation requested
by a task scheduled on that CPU.
But tasks can migrate among cpus, so this definition maybe not accurate.
I also saw some suggestions that it should be a global value, not per
cpu, but don`t see clear definition, so anyone have suggestions about
it? (personally guessing, maybe: there is task waiting for outstanding
I/O while all cpus are idle? But I don't think it is good...)
--
Yours Sincerely,
Cao jin
Peter Zijlstra
Jun 7, 2016, 8:10:06 AM6/7/16
to
per-cpu iowait numbers are doomed.
But I don't see it matters one way or the other. Personally I would just
make the kernel return 0 and be done with it.
Cao jin
Jun 8, 2016, 3:10:06 AM6/8/16
to
Cao jin
Jun 27, 2016, 2:10:05 AM6/27/16
to
Hi Peter
Sorry to bother you on this topic again.
On 06/07/2016 08:01 PM, Peter Zijlstra wrote:
So I just want to make sure, is it acceptable if I send a patch make
iowait a constant 0? Because make it zero seems we are abandoning it(0
doesn'`t have any meanings for it). While maybe some user space tools
monitors this value.
Sorry to bother you on this topic again.
On 06/07/2016 08:01 PM, Peter Zijlstra wrote:
iowait a constant 0? Because make it zero seems we are abandoning it(0
doesn'`t have any meanings for it). While maybe some user space tools
monitors this value.
0 new messages