linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Josef Bacik <josef@toxicpanda.com>
To: linux-btrfs@vger.kernel.org, kernel-team@fb.com
Cc: Zygo Blaxell <ce3g8jdj@umail.furryterror.org>
Subject: [PATCH 01/13] btrfs: use a stable rolling avg for delayed refs avg
Date: Fri, 13 Mar 2020 17:23:18 -0400	[thread overview]
Message-ID: <20200313212330.149024-2-josef@toxicpanda.com> (raw)
In-Reply-To: <20200313212330.149024-1-josef@toxicpanda.com>

From: Zygo Blaxell <ce3g8jdj@umail.furryterror.org>

The way we currently calculate the average of delayed refs is very
jittery.  We do a weighted average, weighing the new average at 3/4 of
the previously calculated average.

This jitteriness leads to pretty wild swings in latencies when we are
generating a lot of delayed refs.  Fix this by smoothing out the delayed
ref average calculation with 1000 seconds of data, 1000 refs minimum,
with a 0.75 decay rate.

Signed-off-by: Zygo Blaxell <ce3g8jdj@umail.furryterror.org>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
---
 fs/btrfs/ctree.h       |  7 +++++++
 fs/btrfs/disk-io.c     |  3 +++
 fs/btrfs/extent-tree.c | 19 +++++++++++++++++--
 3 files changed, 27 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 2ccb2a090782..992ce47977b8 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -620,7 +620,14 @@ struct btrfs_fs_info {
 
 	u64 generation;
 	u64 last_trans_committed;
+
+	/*
+	 * This is for keeping track of how long it takes to run delayed refs so
+	 * that our delayed ref timing doesn't hurt us.
+	 */
 	u64 avg_delayed_ref_runtime;
+	u64 delayed_ref_runtime;
+	u64 delayed_ref_nr_run;
 
 	/*
 	 * this is updated to the current trans every time a full commit
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 772cf0fa7c55..b5846552666e 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2734,6 +2734,9 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info)
 	fs_info->tree_mod_log = RB_ROOT;
 	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 	fs_info->avg_delayed_ref_runtime = NSEC_PER_SEC >> 6; /* div by 64 */
+	fs_info->delayed_ref_runtime = NSEC_PER_SEC;
+	fs_info->delayed_ref_nr_run = 64;
+
 	/* readahead state */
 	INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_DIRECT_RECLAIM);
 	spin_lock_init(&fs_info->reada_lock);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 2925b3ad77a1..645ae95f465e 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -2082,8 +2082,23 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 		 * to avoid large swings in the average.
 		 */
 		spin_lock(&delayed_refs->lock);
-		avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
-		fs_info->avg_delayed_ref_runtime = avg >> 2;	/* div by 4 */
+		fs_info->delayed_ref_nr_run += actual_count;
+		fs_info->delayed_ref_runtime += runtime;
+		avg = div64_u64(fs_info->delayed_ref_runtime,
+				fs_info->delayed_ref_nr_run);
+
+		/*
+		 * Once we've built up a fair bit of data, start decaying
+		 * everything by 3/4.
+		 */
+		if (fs_info->delayed_ref_runtime >= (NSEC_PER_SEC * 1000ULL) &&
+		    fs_info->delayed_ref_nr_run > 1000) {
+			fs_info->delayed_ref_runtime *= 3;
+			fs_info->delayed_ref_runtime >>= 2;
+			fs_info->delayed_ref_nr_run *= 3;
+			fs_info->delayed_ref_nr_run >>= 2;
+		}
+		fs_info->avg_delayed_ref_runtime = avg;
 		spin_unlock(&delayed_refs->lock);
 	}
 	return 0;
-- 
2.24.1


  reply	other threads:[~2020-03-13 21:23 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-13 21:23 [PATCH 00/13] Throttle delayed refs based on time Josef Bacik
2020-03-13 21:23 ` Josef Bacik [this message]
2020-03-13 21:23 ` [PATCH 02/13] btrfs: change btrfs_should_throttle_delayed_refs to a bool Josef Bacik
2020-03-13 21:23 ` [PATCH 03/13] btrfs: make btrfs_should_throttle_delayed_refs only check run time Josef Bacik
2020-03-13 21:23 ` [PATCH 04/13] btrfs: make should_end_transaction check time to run delayed refs Josef Bacik
2020-03-13 21:23 ` [PATCH 05/13] btrfs: squash should_end_transaction Josef Bacik
2020-03-13 21:23 ` [PATCH 06/13] btrfs: add a mode for delayed ref time based throttling Josef Bacik
2020-03-13 21:23 ` [PATCH 07/13] btrfs: kick off async delayed ref flushing if we are over time budget Josef Bacik
2020-04-09 13:11   ` Nikolay Borisov
2020-04-09 13:26   ` Nikolay Borisov
2020-03-13 21:23 ` [PATCH 08/13] btrfs: adjust the arguments for btrfs_should_throttle_delayed_refs Josef Bacik
2020-03-13 21:23 ` [PATCH 09/13] btrfs: throttle delayed refs based on time Josef Bacik
2020-03-13 21:23 ` [PATCH 10/13] btrfs: handle uncontrolled delayed ref generation Josef Bacik
2020-03-13 21:23 ` [PATCH 11/13] btrfs: check delayed refs while relocating Josef Bacik
2020-03-13 21:23 ` [PATCH 12/13] btrfs: throttle truncate for delayed ref generation Josef Bacik
2020-03-13 21:23 ` [PATCH 13/13] btrfs: throttle snapshot delete on delayed refs Josef Bacik

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200313212330.149024-2-josef@toxicpanda.com \
    --to=josef@toxicpanda.com \
    --cc=ce3g8jdj@umail.furryterror.org \
    --cc=kernel-team@fb.com \
    --cc=linux-btrfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).