From d86a238f29c1519bad37bd38d12227bd69d1947f Mon Sep 17 00:00:00 2001 From: Krutika Dhananjay Date: Mon, 8 Jun 2015 11:36:12 +0530 Subject: cluster/afr: Do not attempt entry self-heal if the last lookup on entry failed on src Backport of: http://review.gluster.org/11119 Test bug-948686.t was causing shd to dump core due to gfid being NULL. This was due to the volume being stopped while index heal's in progress, causing afr_selfheal_unlocked_lookup_on() to fail sometimes on the src brick with ENOTCONN. And when afr_selfheal_newentry_mark() copies the gfid off the src iatt, it essentially copies null gfid. This was causing the assertion as part of xattrop in protocol/client to fail. Change-Id: I81723567af824ce4a9fa37e309eeeab8404ac71e BUG: 1233036 Signed-off-by: Krutika Dhananjay Reviewed-on: http://review.gluster.org/11309 Tested-by: Gluster Build System Reviewed-by: Pranith Kumar Karampuri Reviewed-by: Raghavendra Bhat --- xlators/cluster/afr/src/afr-self-heal-entry.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/xlators/cluster/afr/src/afr-self-heal-entry.c b/xlators/cluster/afr/src/afr-self-heal-entry.c index 13825e1eba..7f5095e2ad 100644 --- a/xlators/cluster/afr/src/afr-self-heal-entry.c +++ b/xlators/cluster/afr/src/afr-self-heal-entry.c @@ -178,6 +178,13 @@ __afr_selfheal_heal_dirent (call_frame_t *frame, xlator_t *this, fd_t *fd, if (!replies[source].valid) return -EIO; + /* Skip healing this entry if the last lookup on it failed for reasons + * other than ENOENT. + */ + if ((replies[source].op_ret < 0) && + (replies[source].op_errno != ENOENT)) + return -replies[source].op_errno; + for (i = 0; i < priv->child_count; i++) { if (!healed_sinks[i]) continue; @@ -193,7 +200,7 @@ __afr_selfheal_heal_dirent (call_frame_t *frame, xlator_t *this, fd_t *fd, ret = afr_selfheal_recreate_entry (this, i, source, fd->inode, name, inode, replies, newentry); - } + } if (ret < 0) break; } -- cgit