Bring per_txg_dirty_frees_percent back to 30

The current value causes significant artificial slowdown during mass
parallel file removal, which can be observed both on FreeBSD and Linux
when running real workloads.

Sample results from Linux doing make -j 96 clean after an allyesconfig
modules build:

before: 4.14s user 6.79s system 48% cpu 22.631 total
after:	4.17s user 6.44s system 153% cpu 6.927 total

FreeBSD results in the ticket.

Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Richard Yao <richard.yao@alumni.stonybrook.edu>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by:	Mateusz Guzik <mjguzik@gmail.com>
Closes #13932
Closes #13938

diff --git a/man/man4/zfs.4 b/man/man4/zfs.4
index fcb97d716d542ef3bea67ca7f1fc814bc7e34d6d..b317941bbd362e79ed1998a82a1975017891e848 100644 (file)
--- a/man/man4/zfs.4
+++ b/man/man4/zfs.4
@@ -1651,7 +1651,7 @@ prefetched during a pool traversal, like
 .Nm zfs Cm send
 or other data crawling operations.
 .
-.It Sy zfs_per_txg_dirty_frees_percent Ns = Ns Sy 5 Ns % Pq ulong
+.It Sy zfs_per_txg_dirty_frees_percent Ns = Ns Sy 30 Ns % Pq ulong
 Control percentage of dirtied indirect blocks from frees allowed into one TXG.
 After this threshold is crossed, additional frees will wait until the next TXG.
 .Sy 0 No disables this throttle.

diff --git a/module/zfs/dmu.c b/module/zfs/dmu.c
index 0832fbcfdf40848fd49794485ebe721fc6a51f0c..c67a36470e3382ee7a5f9a541afb690ae76e7774 100644 (file)
--- a/module/zfs/dmu.c
+++ b/module/zfs/dmu.c
@@ -71,7 +71,7 @@ int zfs_nopwrite_enabled = 1;
  * will wait until the next TXG.
  * A value of zero will disable this throttle.
  */
-unsigned long zfs_per_txg_dirty_frees_percent = 5;
+unsigned long zfs_per_txg_dirty_frees_percent = 30;
 
 /*
  * Enable/disable forcing txg sync when dirty checking for holes with lseek().