fio: reset more counters when ramp time has elapsed

From: Vincent Fu <[email protected]>

Jens, I sent the wrong patch earlier.

This version should be safer since reset_io_stats() is now called in ramp_time_over() instead of in reset_all_stats().

Vincent Fu (1):
fio: reset more counters when ramp time has elapsed

 time.c | 1 +
 1 file changed, 1 insertion(+)

2.17.1

From: Vincent Fu <[email protected]>

Currently,

fio --name=test --ramp_time=2 --time_based=1 --runtime=2s --size=1M

will produce an output line like

IO depths : 1=184.7%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%

184.7% appears because the denominator used in the IO depth calculation is reset at the conclusion of ramp_time but the numerator is not. Latency and other IO depth statistics also include data from IOs completed during ramp time. This patch resets these counters when ramp time is over.

---
 time.c | 1 +
 1 file changed, 1 insertion(+)

index c8876829..19999699 100644
@@ -118,6 +118,7 @@ bool ramp_time_over(struct thread_data *td)
 	if (utime_since_now(&td->epoch) >= td->o.ramp_time) {
 		td->ramp_time_over = true;
 		reset_all_stats(td);
+		reset_io_stats(td);
 		td_set_runstate(td, TD_RAMP);
 
 		/*

2.17.1

I reproduced this with fio d5dbacf662b1cc3fb09b5cd70b236ab98d1c0dbe :

[sboisvert@GT480:fio]$ ./fio --name=test --ramp_time=2 --time_based=1 --runtime=2s --size=1M | grep "IO depths"
  IO depths    : 1=184.1%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%

[sboisvert@GT480:fio]$ ./fio --name=test --ramp_time=2 --time_based=1 --runtime=2s --size=1M | grep "IO depths"
  IO depths    : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%

Agree, that makes me less nervous. Applied, thanks.

Jens Axboe