Adaptive io latency distrobution "histogram"
Log ranges all the way down to microseconds, and display the most
appropriate values for those ranges.
This makes the latency usable for sub-milisecond devices as well.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
diff --git a/fio.h b/fio.h
index 2104e3a..813e55b 100644
--- a/fio.h
+++ b/fio.h
@@ -295,7 +295,8 @@
* How many depth levels to log
*/
#define FIO_IO_U_MAP_NR 8
-#define FIO_IO_U_LAT_NR 12
+#define FIO_IO_U_LAT_U_NR 12
+#define FIO_IO_U_LAT_M_NR 10
struct thread_stat {
char *name;
@@ -333,7 +334,8 @@
* IO depth and latency stats
*/
unsigned int io_u_map[FIO_IO_U_MAP_NR];
- unsigned int io_u_lat[FIO_IO_U_LAT_NR];
+ unsigned int io_u_lat_u[FIO_IO_U_LAT_U_NR];
+ unsigned int io_u_lat_m[FIO_IO_U_LAT_M_NR];
unsigned long total_io_u[2];
unsigned long short_io_u[2];
diff --git a/io_u.c b/io_u.c
index 9aab97f..e0614bd 100644
--- a/io_u.c
+++ b/io_u.c
@@ -387,38 +387,98 @@
td->ts.total_io_u[io_u->ddir]++;
}
-static void io_u_mark_latency(struct thread_data *td, unsigned long msec)
+static void io_u_mark_lat_usec(struct thread_data *td, unsigned long usec)
+{
+ int index = 0;
+
+ assert(usec < 1000);
+
+ switch (usec) {
+ case 750 ... 999:
+ index = 9;
+ break;
+ case 500 ... 749:
+ index = 8;
+ break;
+ case 250 ... 499:
+ index = 7;
+ break;
+ case 100 ... 249:
+ index = 6;
+ break;
+ case 50 ... 99:
+ index = 5;
+ break;
+ case 20 ... 49:
+ index = 4;
+ break;
+ case 10 ... 19:
+ index = 3;
+ break;
+ case 4 ... 9:
+ index = 2;
+ break;
+ case 2 ... 3:
+ index = 1;
+ case 0 ... 1:
+ break;
+ }
+
+ assert(index < FIO_IO_U_LAT_U_NR);
+ td->ts.io_u_lat_u[index]++;
+}
+
+static void io_u_mark_lat_msec(struct thread_data *td, unsigned long msec)
{
int index = 0;
switch (msec) {
default:
- index++;
+ index = 11;
+ break;
case 1000 ... 1999:
- index++;
+ index = 10;
+ break;
case 750 ... 999:
- index++;
+ index = 9;
+ break;
case 500 ... 749:
- index++;
+ index = 8;
+ break;
case 250 ... 499:
- index++;
+ index = 7;
+ break;
case 100 ... 249:
- index++;
+ index = 6;
+ break;
case 50 ... 99:
- index++;
+ index = 5;
+ break;
case 20 ... 49:
- index++;
+ index = 4;
+ break;
case 10 ... 19:
- index++;
+ index = 3;
+ break;
case 4 ... 9:
- index++;
+ index = 2;
+ break;
case 2 ... 3:
- index++;
+ index = 1;
case 0 ... 1:
break;
}
- td->ts.io_u_lat[index]++;
+ assert(index < FIO_IO_U_LAT_M_NR);
+ td->ts.io_u_lat_m[index]++;
+}
+
+static void io_u_mark_latency(struct thread_data *td, unsigned long usec)
+{
+ if (usec < 1000)
+ io_u_mark_lat_usec(td, usec);
+ else
+ io_u_mark_lat_msec(td, usec / 1000);
}
/*
@@ -682,7 +742,7 @@
add_clat_sample(td, idx, usec);
add_bw_sample(td, idx, &icd->time);
- io_u_mark_latency(td, usec / 1000);
+ io_u_mark_latency(td, usec);
if (td_write(td) && idx == DDIR_WRITE &&
td->o.verify != VERIFY_NONE)
diff --git a/stat.c b/stat.c
index 34c8fce..28d85c1 100644
--- a/stat.c
+++ b/stat.c
@@ -131,21 +131,32 @@
}
}
-static void stat_calc_lat(struct thread_stat *ts, double *io_u_lat)
+static void stat_calc_lat(struct thread_stat *ts, double *dst,
+ unsigned int *src, int nr)
{
int i;
/*
* Do latency distribution calculations
*/
- for (i = 0; i < FIO_IO_U_LAT_NR; i++) {
- io_u_lat[i] = (double) ts->io_u_lat[i] / (double) ts_total_io_u(ts);
- io_u_lat[i] *= 100.0;
- if (io_u_lat[i] < 0.01 && ts->io_u_lat[i])
- io_u_lat[i] = 0.01;
+ for (i = 0; i < nr; i++) {
+ dst[i] = (double) src[i] / (double) ts_total_io_u(ts);
+ dst[i] *= 100.0;
+ if (dst[i] < 0.01 && src[i])
+ dst[i] = 0.01;
}
}
+static void stat_calc_lat_u(struct thread_stat *ts, double *io_u_lat)
+{
+ stat_calc_lat(ts, io_u_lat, ts->io_u_lat_u, FIO_IO_U_LAT_U_NR);
+}
+
+static void stat_calc_lat_m(struct thread_stat *ts, double *io_u_lat)
+{
+ stat_calc_lat(ts, io_u_lat, ts->io_u_lat_m, FIO_IO_U_LAT_M_NR);
+}
+
static int usec_to_msec(unsigned long *min, unsigned long *max, double *mean,
double *dev)
{
@@ -208,13 +219,61 @@
}
}
+static void show_lat(double *io_u_lat, int nr, const char **ranges,
+ const char *msg)
+{
+ int new_line = 1, i, line = 0;
+
+ for (i = 0; i < nr; i++) {
+ if (io_u_lat[i] <= 0.0)
+ continue;
+ if (new_line) {
+ log_info(" lat (%s): ", msg);
+ new_line = 0;
+ line = 0;
+ }
+ if (line)
+ log_info(", ");
+ log_info("%s%3.2f%%", ranges[i], io_u_lat[i]);
+ line++;
+ if (line == 5)
+ new_line = 1;
+ }
+
+}
+
+static void show_lat_u(double *io_u_lat_u)
+{
+ const char *ranges[] = { "2=", "4=", "10=", "20=", "50=", "100=",
+ "250=", "500=", "750=", "1000=", };
+
+ show_lat(io_u_lat_u, FIO_IO_U_LAT_U_NR, ranges, "usec");
+}
+
+static void show_lat_m(double *io_u_lat_m)
+{
+ const char *ranges[] = { "2=", "4=", "10=", "20=", "50=", "100=",
+ "250=", "500=", "750=", "1000=", "2000=",
+ ">=2000=", };
+
+ show_lat(io_u_lat_m, FIO_IO_U_LAT_M_NR, ranges, "msec");
+}
+
+static void show_latencies(double *io_u_lat_u, double *io_u_lat_m)
+{
+ show_lat_u(io_u_lat_u);
+ show_lat_m(io_u_lat_m);
+ log_info("\n");
+}
+
static void show_thread_status(struct thread_stat *ts,
struct group_run_stats *rs)
{
double usr_cpu, sys_cpu;
unsigned long runtime;
double io_u_dist[FIO_IO_U_MAP_NR];
- double io_u_lat[FIO_IO_U_LAT_NR];
+ double io_u_lat_u[FIO_IO_U_LAT_U_NR];
+ double io_u_lat_m[FIO_IO_U_LAT_M_NR];
if (!(ts->io_bytes[0] + ts->io_bytes[1]))
return;
@@ -246,14 +305,13 @@
log_info(" cpu : usr=%3.2f%%, sys=%3.2f%%, ctx=%lu\n", usr_cpu, sys_cpu, ts->ctx);
stat_calc_dist(ts, io_u_dist);
- stat_calc_lat(ts, io_u_lat);
+ stat_calc_lat_u(ts, io_u_lat_u);
+ stat_calc_lat_m(ts, io_u_lat_m);
log_info(" IO depths : 1=%3.1f%%, 2=%3.1f%%, 4=%3.1f%%, 8=%3.1f%%, 16=%3.1f%%, 32=%3.1f%%, >=64=%3.1f%%\n", io_u_dist[0], io_u_dist[1], io_u_dist[2], io_u_dist[3], io_u_dist[4], io_u_dist[5], io_u_dist[6]);
log_info(" issued r/w: total=%lu/%lu, short=%lu/%lu\n", ts->total_io_u[0], ts->total_io_u[1], ts->short_io_u[0], ts->short_io_u[1]);
- log_info(" lat (msec): 2=%3.2f%%, 4=%3.2f%%, 10=%3.2f%%, 20=%3.2f%%, 50=%3.2f%%\n", io_u_lat[0], io_u_lat[1], io_u_lat[2], io_u_lat[3], io_u_lat[4]);
- log_info(" lat (msec): 100=%3.2f%%, 250=%3.2f%%, 500=%3.2f%%, 750=%3.2f%%\n", io_u_lat[5], io_u_lat[6], io_u_lat[7], io_u_lat[8]);
- log_info(" lat (msec): 1000=%3.2f%%, 2000=%3.2f%%, >=2000=%3.2f%%\n", io_u_lat[9], io_u_lat[10], io_u_lat[11]);
+ show_latencies(io_u_lat_u, io_u_lat_m);
}
static void show_ddir_status_terse(struct thread_stat *ts,
@@ -293,8 +351,10 @@
struct group_run_stats *rs)
{
double io_u_dist[FIO_IO_U_MAP_NR];
- double io_u_lat[FIO_IO_U_LAT_NR];
+ double io_u_lat_u[FIO_IO_U_LAT_U_NR];
+ double io_u_lat_m[FIO_IO_U_LAT_M_NR];
double usr_cpu, sys_cpu;
+ int i;
log_info("%s;%d;%d", ts->name, ts->groupid, ts->error);
@@ -314,12 +374,16 @@
log_info(";%f%%;%f%%;%lu", usr_cpu, sys_cpu, ts->ctx);
stat_calc_dist(ts, io_u_dist);
- stat_calc_lat(ts, io_u_lat);
+ stat_calc_lat_u(ts, io_u_lat_u);
+ stat_calc_lat_m(ts, io_u_lat_m);
log_info(";%3.1f%%;%3.1f%%;%3.1f%%;%3.1f%%;%3.1f%%;%3.1f%%;%3.1f%%", io_u_dist[0], io_u_dist[1], io_u_dist[2], io_u_dist[3], io_u_dist[4], io_u_dist[5], io_u_dist[6]);
- log_info(";%3.2f%%;%3.2f%%;%3.2f%%;%3.2f%%;%3.2f%%;%3.2f%%\n", io_u_lat[0], io_u_lat[1], io_u_lat[2], io_u_lat[3], io_u_lat[4], io_u_lat[5]);
- log_info(";%3.2f%%;%3.2f%%;%3.2f%%;%3.2f%%;%3.2f%%", io_u_lat[6], io_u_lat[7], io_u_lat[8], io_u_lat[9], io_u_lat[10]);
+ for (i = 0; i < FIO_IO_U_LAT_U_NR; i++)
+ log_info(";%3.2f%%", io_u_lat_u[i]);
+ for (i = 0; i < FIO_IO_U_LAT_M_NR; i++)
+ log_info(";%3.2f%%", io_u_lat_m[i]);
+ log_info("\n");
if (ts->description)
log_info(";%s", ts->description);
@@ -454,8 +518,11 @@
for (k = 0; k < FIO_IO_U_MAP_NR; k++)
ts->io_u_map[k] += td->ts.io_u_map[k];
- for (k = 0; k < FIO_IO_U_LAT_NR; k++)
- ts->io_u_lat[k] += td->ts.io_u_lat[k];
+ for (k = 0; k < FIO_IO_U_LAT_U_NR; k++)
+ ts->io_u_lat_u[k] += td->ts.io_u_lat_u[k];
+ for (k = 0; k < FIO_IO_U_LAT_M_NR; k++)
+ ts->io_u_lat_m[k] += td->ts.io_u_lat_m[k];
+
for (k = 0; k <= DDIR_WRITE; k++) {
ts->total_io_u[k] += td->ts.total_io_u[k];