diff options
author | Michael Lyle <mlyle@lyle.org> | 2017-10-13 16:35:37 -0700 |
---|---|---|
committer | Jens Axboe <axboe@kernel.dk> | 2017-10-16 09:07:26 -0600 |
commit | ae82ddbfeb359fcffa97be5fb5bcd59165f2864f (patch) | |
tree | 6aa8da536120096b77cd15befe59d9b3457adb71 /drivers/md/bcache/util.c | |
parent | 1d316e658374f700fdfff9299c70ce65d8d145e6 (diff) |
bcache: smooth writeback rate control
This works in conjunction with the new PI controller. Currently, in
real-world workloads, the rate controller attempts to write back 1
sector per second. In practice, these minimum-rate writebacks are
between 4k and 60k in test scenarios, since bcache aggregates and
attempts to do contiguous writes and because filesystems on top of
bcachefs typically write 4k or more.
Previously, bcache used to guarantee to write at least once per second.
This means that the actual writeback rate would exceed the configured
amount by a factor of 8-120 or more.
This patch adjusts to be willing to sleep up to 2.5 seconds, and to
target writing 4k/second. On the smallest writes, it will sleep 1
second like before, but many times it will sleep longer and load the
backing device less. This keeps the loading on the cache and backing
device related to writeback more consistent when writing back at low
rates.
Signed-off-by: Michael Lyle <mlyle@lyle.org>
Reviewed-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'drivers/md/bcache/util.c')
-rw-r--r-- | drivers/md/bcache/util.c | 10 |
1 files changed, 8 insertions, 2 deletions
diff --git a/drivers/md/bcache/util.c b/drivers/md/bcache/util.c index 176d3c2ef5f5..4dbe37e82877 100644 --- a/drivers/md/bcache/util.c +++ b/drivers/md/bcache/util.c @@ -232,8 +232,14 @@ uint64_t bch_next_delay(struct bch_ratelimit *d, uint64_t done) d->next += div_u64(done * NSEC_PER_SEC, d->rate); - if (time_before64(now + NSEC_PER_SEC, d->next)) - d->next = now + NSEC_PER_SEC; + /* Bound the time. Don't let us fall further than 2 seconds behind + * (this prevents unnecessary backlog that would make it impossible + * to catch up). If we're ahead of the desired writeback rate, + * don't let us sleep more than 2.5 seconds (so we can notice/respond + * if the control system tells us to speed up!). + */ + if (time_before64(now + NSEC_PER_SEC * 5 / 2, d->next)) + d->next = now + NSEC_PER_SEC * 5 / 2; if (time_after64(now - NSEC_PER_SEC * 2, d->next)) d->next = now - NSEC_PER_SEC * 2; |