Blob Blame History Raw
From b52044163b8e971b1f9c2e7bd2d78a9dfcce1fe0 Mon Sep 17 00:00:00 2001
From: Pranith Kumar K <pkarampu@redhat.com>
Date: Tue, 7 Mar 2017 14:32:18 +0530
Subject: [PATCH 303/303] cluster/ec: Don't mark dirty on entry/meta ops in
 query-info

	Backport of https://review.gluster.org/16865

We wanted to mark dirty for metadata/entry operations
whenever query-info is set and info is not yet there because we
are anyway sending xattrop over the network. But this is causing
25% regression from 3.8.8 so removing this optimization

BUG: 1408655
Change-Id: Iabb0b64bd4d3623688790e4b67e5c20b4da977a1
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: https://code.engineering.redhat.com/gerrit/99547
---
 xlators/cluster/ec/src/ec-common.c | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/xlators/cluster/ec/src/ec-common.c b/xlators/cluster/ec/src/ec-common.c
index 647e750..f54084d 100644
--- a/xlators/cluster/ec/src/ec-common.c
+++ b/xlators/cluster/ec/src/ec-common.c
@@ -1130,12 +1130,6 @@ void ec_get_size_version(ec_lock_link_t *link)
         !(ec->node_mask & ~link->lock->good_mask) && !ec_is_data_fop (fop->id))
             link->optimistic_changelog = _gf_true;
 
-    /* If ctx->have_info is false and lock->query is true, it means that we'll
-     * send the xattrop anyway, so we can use it to update dirty counts, even
-     * if it's not necessary to do it right now. */
-    if (!ctx->have_info && lock->query)
-            link->optimistic_changelog = _gf_false;
-
     set_dirty = ec_set_dirty_flag (link, ctx, dirty);
 
     /* If ec metadata has already been retrieved, do not try again. */
-- 
2.9.3