glusterfs/0259-gluster-block-tuning-perf-options.patch

48 lines
1.6 KiB
Diff
Raw Normal View History

From 775a62906030e5b5dc60f17284a7d516ce4118f9 Mon Sep 17 00:00:00 2001
From: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
Date: Thu, 27 Jun 2019 13:18:32 +0530
Subject: [PATCH 259/261] gluster-block: tuning perf options
As per the perf experiment run by Elvir, with NVME devices used for OCP (gluster)
it was seen that particularly read operations (read/randread) benefited from
these options.
> upstream patch : https://review.gluster.org/#/c/glusterfs/+/22963/
>Change-Id: Ibec4b96afd28e6f7e757b6ef203ccdbc0d9854d5
>Fixes: bz#1727852
>Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
Change-Id: Ibec4b96afd28e6f7e757b6ef203ccdbc0d9854d5
BUG: 1708180
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
Reviewed-on: https://code.engineering.redhat.com/gerrit/176724
Tested-by: RHGS Build Bot <nigelb@redhat.com>
Reviewed-by: Atin Mukherjee <amukherj@redhat.com>
---
extras/group-gluster-block | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/extras/group-gluster-block b/extras/group-gluster-block
index 56b406e..1e39801 100644
--- a/extras/group-gluster-block
+++ b/extras/group-gluster-block
@@ -5,6 +5,14 @@ performance.stat-prefetch=off
performance.open-behind=off
performance.readdir-ahead=off
performance.strict-o-direct=on
+performance.client-io-threads=on
+performance.io-thread-count=32
+performance.high-prio-threads=32
+performance.normal-prio-threads=32
+performance.low-prio-threads=32
+performance.least-prio-threads=4
+client.event-threads=8
+server.event-threads=8
network.remote-dio=disable
cluster.eager-lock=enable
cluster.quorum-type=auto
--
1.8.3.1