* QEMU throttling infrastructure
*
* Copyright (C) Nodalink, EURL. 2013-2014
- * Copyright (C) Igalia, S.L. 2015
+ * Copyright (C) Igalia, S.L. 2015-2016
*
* Authors:
* BenoƮt Canet <benoit.canet@nodalink.com>
} BucketType;
/*
- * The max parameter of the leaky bucket throttling algorithm can be used to
- * allow the guest to do bursts.
- * The max value is a pool of I/O that the guest can use without being throttled
- * at all. Throttling is triggered once this pool is empty.
+ * This module implements I/O limits using the leaky bucket
+ * algorithm. The code is independent of the I/O units, but it is
+ * currently used for bytes per second and operations per second.
+ *
+ * Three parameters can be set by the user:
+ *
+ * - avg: the desired I/O limits in units per second.
+ * - max: the limit during bursts, also in units per second.
+ * - burst_length: the maximum length of the burst period, in seconds.
+ *
+ * Here's how it works:
+ *
+ * - The bucket level (number of performed I/O units) is kept in
+ * bkt.level and leaks at a rate of bkt.avg units per second.
+ *
+ * - The size of the bucket is bkt.max * bkt.burst_length. Once the
+ * bucket is full no more I/O is performed until the bucket leaks
+ * again. This is what makes the I/O rate bkt.avg.
+ *
+ * - The bkt.avg rate does not apply until the bucket is full,
+ * allowing the user to do bursts until then. The I/O limit during
+ * bursts is bkt.max. To enforce this limit we keep an additional
+ * bucket in bkt.burst_length that leaks at a rate of bkt.max units
+ * per second.
+ *
+ * - Because of all of the above, the user can perform I/O at a
+ * maximum of bkt.max units per second for at most bkt.burst_length
+ * seconds in a row. After that the bucket will be full and the I/O
+ * rate will go down to bkt.avg.
+ *
+ * - Since the bucket always leaks at a rate of bkt.avg, this also
+ * determines how much the user needs to wait before being able to
+ * do bursts again.
*/
typedef struct LeakyBucket {
double avg; /* average goal in units per second */
double max; /* leaky bucket max burst in units */
double level; /* bucket level in units */
+ double burst_level; /* bucket level in units (for computing bursts) */
+ unsigned burst_length; /* max length of the burst period, in seconds */
} LeakyBucket;
/* The following structure is used to configure a ThrottleState
/* make the bucket leak */
bkt->level = MAX(bkt->level - leak, 0);
+
+ /* if we allow bursts for more than one second we also need to
+ * keep track of bkt->burst_level so the bkt->max goal per second
+ * is attained */
+ if (bkt->burst_length > 1) {
+ leak = (bkt->max * (double) delta_ns) / NANOSECONDS_PER_SECOND;
+ bkt->burst_level = MAX(bkt->burst_level - leak, 0);
+ }
}
/* Calculate the time delta since last leak and make proportionals leaks
return 0;
}
- extra = bkt->level - bkt->max;
+ /* If the bucket is full then we have to wait */
+ extra = bkt->level - bkt->max * bkt->burst_length;
+ if (extra > 0) {
+ return throttle_do_compute_wait(bkt->avg, extra);
+ }
- if (extra <= 0) {
- return 0;
+ /* If the bucket is not full yet we have to make sure that we
+ * fulfill the goal of bkt->max units per second. */
+ if (bkt->burst_length > 1) {
+ /* We use 1/10 of the max value to smooth the throttling.
+ * See throttle_fix_bucket() for more details. */
+ extra = bkt->burst_level - bkt->max / 10;
+ if (extra > 0) {
+ return throttle_do_compute_wait(bkt->max, extra);
+ }
}
- return throttle_do_compute_wait(bkt->avg, extra);
+ return 0;
}
/* This function compute the time that must be waited while this IO
*/
void throttle_config_init(ThrottleConfig *cfg)
{
+ unsigned i;
memset(cfg, 0, sizeof(*cfg));
+ for (i = 0; i < BUCKETS_COUNT; i++) {
+ cfg->buckets[i].burst_length = 1;
+ }
}
/* To be called first on the ThrottleState */
return false;
}
+ if (!cfg->buckets[i].burst_length) {
+ error_setg(errp, "the burst length cannot be 0");
+ return false;
+ }
+
+ if (cfg->buckets[i].burst_length > 1 && !cfg->buckets[i].max) {
+ error_setg(errp, "burst length set without burst rate");
+ return false;
+ }
+
if (cfg->buckets[i].max && !cfg->buckets[i].avg) {
error_setg(errp, "bps_max/iops_max require corresponding"
" bps/iops values");
double min;
/* zero bucket level */
- bkt->level = 0;
+ bkt->level = bkt->burst_level = 0;
/* The following is done to cope with the Linux CFQ block scheduler
* which regroup reads and writes by block of 100ms in the guest.
*/
void throttle_account(ThrottleState *ts, bool is_write, uint64_t size)
{
+ const BucketType bucket_types_size[2][2] = {
+ { THROTTLE_BPS_TOTAL, THROTTLE_BPS_READ },
+ { THROTTLE_BPS_TOTAL, THROTTLE_BPS_WRITE }
+ };
+ const BucketType bucket_types_units[2][2] = {
+ { THROTTLE_OPS_TOTAL, THROTTLE_OPS_READ },
+ { THROTTLE_OPS_TOTAL, THROTTLE_OPS_WRITE }
+ };
double units = 1.0;
+ unsigned i;
/* if cfg.op_size is defined and smaller than size we compute unit count */
if (ts->cfg.op_size && size > ts->cfg.op_size) {
units = (double) size / ts->cfg.op_size;
}
- ts->cfg.buckets[THROTTLE_BPS_TOTAL].level += size;
- ts->cfg.buckets[THROTTLE_OPS_TOTAL].level += units;
+ for (i = 0; i < 2; i++) {
+ LeakyBucket *bkt;
+
+ bkt = &ts->cfg.buckets[bucket_types_size[is_write][i]];
+ bkt->level += size;
+ if (bkt->burst_length > 1) {
+ bkt->burst_level += size;
+ }
- if (is_write) {
- ts->cfg.buckets[THROTTLE_BPS_WRITE].level += size;
- ts->cfg.buckets[THROTTLE_OPS_WRITE].level += units;
- } else {
- ts->cfg.buckets[THROTTLE_BPS_READ].level += size;
- ts->cfg.buckets[THROTTLE_OPS_READ].level += units;
+ bkt = &ts->cfg.buckets[bucket_types_units[is_write][i]];
+ bkt->level += units;
+ if (bkt->burst_length > 1) {
+ bkt->burst_level += units;
+ }
}
}