Skip to contentSkip to conversation

iocost: better trace vrate changes

PATCH In review Tejun Heo 2 participants 4 messages Sep 25, 2019 – Sep 26, 2019

Subsystems (3)
lore
Share & tools
lore

Tejun Heo

author maintainer
#0 lore raw

PATCH 1/3 iocost: better trace vrate changes

vrate_adj tracepoint traces vrate changes; however, it does so only when busy_level is non-zero. busy_level turning to zero can sometimes be as interesting an event. This patch also enables vrate_adj tracepoint on other vrate related events - busy_level changes and non-zero nr_lagging.

---
Hello, Jens.

I've encountered vrate regulation issues while testing on a hard disk machine. These are three patches to improve vrate adj visibility and fix the issue.

Thanks.

 block/blk-iocost.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
block/blk-iocost.c +6 -1
@@ -1343,7 +1343,7 @@ static void ioc_timer_fn(struct timer_li
 	u32 ppm_wthr = MILLION - ioc->params.qos[QOS_WPPM];
 	u32 missed_ppm[2], rq_wait_pct;
 	u64 period_vtime;
-	int i;
+	int prev_busy_level, i;
 
 	/* how were the latencies during the period? */
 	ioc_lat_stat(ioc, missed_ppm, &rq_wait_pct);
@@ -1531,6 +1531,7 @@ skip_surplus_transfers:
 	 * and experiencing shortages but not surpluses, we're too stingy
 	 * and should increase vtime rate.
 	 */
+	prev_busy_level = ioc->busy_level;
 	if (rq_wait_pct > RQ_WAIT_BUSY_PCT ||
 	    missed_ppm[READ] > ppm_rthr ||
 	    missed_ppm[WRITE] > ppm_wthr) {
@@ -1592,6 +1593,10 @@ skip_surplus_transfers:
 		atomic64_set(&ioc->vtime_rate, vrate);
 		ioc->inuse_margin_vtime = DIV64_U64_ROUND_UP(
 			ioc->period_us * vrate * INUSE_MARGIN_PCT, 100);
+	} else if (ioc->busy_level != prev_busy_level || nr_lagging) {
+		trace_iocost_ioc_vrate_adj(ioc, atomic64_read(&ioc->vtime_rate),
+					   &missed_ppm, rq_wait_pct, nr_lagging,
+					   nr_shortages, nr_surpluses);
 	}
 
 	ioc_refresh_params(ioc, false);
author maintainer
#1 lore raw

PATCH 2/3 iocost: improve nr_lagging handling

Some IOs may span multiple periods. As latencies are collected on completion, the inbetween periods won't register them and may incorrectly decide to increase vrate. nr_lagging tracks these IOs to avoid those situations. Currently, whenever there are IOs which are spanning from the previous period, busy_level is reset to 0 if negative thus suppressing vrate increase.

This has the following two problems.

* When latency target percentiles aren't set, vrate adjustment should only be governed by queue depth depletion; however, the current code keeps nr_lagging active which pulls in latency results and can keep down vrate unexpectedly.

* When lagging condition is detected, it resets the entire negative busy_level. This turned out to be way too aggressive on some devices which sometimes experience extended latencies on a small subset of commands. In addition, a lagging IO will be accounted as latency target miss on completion anyway and resetting busy_level amplifies its impact unnecessarily.

This patch fixes the above two problems by disabling nr_lagging counting when latency target percentiles aren't set and blocking vrate increases when there are lagging IOs while leaving busy_level as-is.

---
 block/blk-iocost.c |   19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)
block/blk-iocost.c +11 -8
@@ -1407,7 +1407,8 @@ static void ioc_timer_fn(struct timer_li
 		 * comparing vdone against period start.  If lagging behind
 		 * IOs from past periods, don't increase vrate.
 		 */
-		if (!atomic_read(&iocg_to_blkg(iocg)->use_delay) &&
+		if ((ppm_rthr != MILLION || ppm_wthr != MILLION) &&
+		    !atomic_read(&iocg_to_blkg(iocg)->use_delay) &&
 		    time_after64(vtime, vdone) &&
 		    time_after64(vtime, now.vnow -
 				 MAX_LAGGING_PERIODS * period_vtime) &&
@@ -1537,21 +1538,23 @@ skip_surplus_transfers:
 	    missed_ppm[WRITE] > ppm_wthr) {
 		ioc->busy_level = max(ioc->busy_level, 0);
 		ioc->busy_level++;
-	} else if (nr_lagging) {
-		ioc->busy_level = max(ioc->busy_level, 0);
-	} else if (nr_shortages && !nr_surpluses &&
-		   rq_wait_pct <= RQ_WAIT_BUSY_PCT * UNBUSY_THR_PCT / 100 &&
+	} else if (rq_wait_pct <= RQ_WAIT_BUSY_PCT * UNBUSY_THR_PCT / 100 &&
 		   missed_ppm[READ] <= ppm_rthr * UNBUSY_THR_PCT / 100 &&
 		   missed_ppm[WRITE] <= ppm_wthr * UNBUSY_THR_PCT / 100) {
-		ioc->busy_level = min(ioc->busy_level, 0);
-		ioc->busy_level--;
+		/* take action iff there is contention */
+		if (nr_shortages && !nr_lagging) {
+			ioc->busy_level = min(ioc->busy_level, 0);
+			/* redistribute surpluses first */
+			if (!nr_surpluses)
+				ioc->busy_level--;
+		}
 	} else {
 		ioc->busy_level = 0;
 	}
 
 	ioc->busy_level = clamp(ioc->busy_level, -1000, 1000);
 
-	if (ioc->busy_level) {
+	if (ioc->busy_level > 0 || (ioc->busy_level < 0 && !nr_lagging)) {
 		u64 vrate = atomic64_read(&ioc->vtime_rate);
 		u64 vrate_min = ioc->vrate_min, vrate_max = ioc->vrate_max;
 
author maintainer
#2 lore raw

PATCH 3/3 iocost: bump up default latency targets for hard disks

The default hard disk param sets latency targets at 50ms. As the default target percentiles are zero, these don't directly regulate vrate; however, they're still used to calculate the period length - 100ms in this case.

This is excessively low. A SATA drive with QD32 saturated with random IOs can easily reach avg completion latency of several hundred msecs. A period duration which is substantially lower than avg completion latency can lead to wildly fluctuating vrate.

Let's bump up the default latency targets to 250ms so that the period duration is sufficiently long.

---
 block/blk-iocost.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
block/blk-iocost.c +2 -2
@@ -529,8 +529,8 @@ struct iocg_wake_ctx {
 static const struct ioc_params autop[] = {
 	[AUTOP_HDD] = {
 		.qos				= {
-			[QOS_RLAT]		=         50000, /* 50ms */
-			[QOS_WLAT]		=         50000,
+			[QOS_RLAT]		=        250000, /* 250ms */
+			[QOS_WLAT]		=        250000,
 			[QOS_MIN]		= VRATE_MIN_PPM,
 			[QOS_MAX]		= VRATE_MAX_PPM,
 		},
maintainer
#3 lore raw
In reply to Tejun Heo
Quoted 6 lines · Tejun Heo

vrate_adj tracepoint traces vrate changes; however, it does so only when busy_level is non-zero. busy_level turning to zero can sometimes be as interesting an event. This patch also enables vrate_adj tracepoint on other vrate related events - busy_level changes and non-zero nr_lagging.

Quoted 7 lines

---
Hello, Jens.

I've encountered vrate regulation issues while testing on a hard disk machine. These are three patches to improve vrate adj visibility and fix the issue.

Applied 1-3 for 5.4, thanks Tejun.

Signature · 1 line
Jens Axboe