From 6e1e90d64cf26d66ff0b5a50fa294be162356ada Mon Sep 17 00:00:00 2001 From: Alexander Motin Date: Wed, 4 May 2022 14:33:42 -0400 Subject: [PATCH] Improve mg_aliquot math When calculating mg_aliquot alike to #12046 use number of unique data disks in the vdev, not the total number of children vdev. Increase default value of the tunable from 512KB to 1MB to compensate. Before this change each disk in striped pool was getting 512KB of sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB. After this change in all the cases each disk should get 1MB. Reviewed-by: Brian Behlendorf Reviewed-by: Ryan Moeller Signed-off-by: Alexander Motin Sponsored-By: iXsystems, Inc. Closes #13388 --- man/man4/zfs.4 | 6 +++--- module/zfs/metaslab.c | 9 +++++---- 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/man/man4/zfs.4 b/man/man4/zfs.4 index 5e4587285b..6921116848 100644 --- a/man/man4/zfs.4 +++ b/man/man4/zfs.4 @@ -213,12 +213,12 @@ For L2ARC devices less than 1GB, the amount of data evicts is significant compared to the amount of restored L2ARC data. In this case, do not write log blocks in L2ARC in order not to waste space. . -.It Sy metaslab_aliquot Ns = Ns Sy 524288 Ns B Po 512kB Pc Pq ulong +.It Sy metaslab_aliquot Ns = Ns Sy 1048576 Ns B Po 1MB Pc Pq ulong Metaslab granularity, in bytes. This is roughly similar to what would be referred to as the "stripe size" in traditional RAID arrays. -In normal operation, ZFS will try to write this amount of data -to a top-level vdev before moving on to the next one. +In normal operation, ZFS will try to write this amount of data to each disk +before moving on to the next top-level vdev. . .It Sy metaslab_bias_enabled Ns = Ns Sy 1 Ns | Ns 0 Pq int Enable metaslab group biasing based on their vdevs' over- or under-utilization diff --git a/module/zfs/metaslab.c b/module/zfs/metaslab.c index 15f545e2d1..ecc70298dc 100644 --- a/module/zfs/metaslab.c +++ b/module/zfs/metaslab.c @@ -48,10 +48,10 @@ /* * Metaslab granularity, in bytes. This is roughly similar to what would be * referred to as the "stripe size" in traditional RAID arrays. In normal - * operation, we will try to write this amount of data to a top-level vdev - * before moving on to the next one. + * operation, we will try to write this amount of data to each disk before + * moving on to the next top-level vdev. */ -unsigned long metaslab_aliquot = 512 << 10; +static unsigned long metaslab_aliquot = 1024 * 1024; /* * For testing, make some blocks above a certain size be gang blocks. @@ -899,7 +899,8 @@ metaslab_group_activate(metaslab_group_t *mg) if (++mg->mg_activation_count <= 0) return; - mg->mg_aliquot = metaslab_aliquot * MAX(1, mg->mg_vd->vdev_children); + mg->mg_aliquot = metaslab_aliquot * MAX(1, + vdev_get_ndisks(mg->mg_vd) - vdev_get_nparity(mg->mg_vd)); metaslab_group_alloc_update(mg); if ((mgprev = mc->mc_allocator[0].mca_rotor) == NULL) {