cluster/afr: Delay post-op for fsync

Problem:
AFR doesn't delay post-op for fsync fop. For fsync heavy workloads
this leads to un-necessary fxattrop/finodelk for every fsync leading
to bad performance.

Fix:
Have delayed post-op for fsync. Add special flag in xdata to indicate
that afr shouldn't delay post-op in cases where either the
process will terminate or graph-switch would happen. Otherwise it leads
to un-necessary heals when the graph-switch/process-termination
happens before delayed-post-op completes.

Fixes: #1253
Change-Id: I531940d13269a111c49e0510d49514dc169f4577
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>

1 files changed, 9 insertions, 2 deletions

diff --git a/xlators/cluster/afr/src/afr-inode-write.c b/xlators/cluster/afr/src/afr-inode-write.c
index c01b4131d58..1d6e4f3570a 100644
--- a/xlators/cluster/afr/src/afr-inode-write.c
+++ b/xlators/cluster/afr/src/afr-inode-write.c
@@ -2506,6 +2506,7 @@ afr_fsync(call_frame_t *frame, xlator_t *this, fd_t *fd, int32_t datasync,
     call_frame_t *transaction_frame = NULL;
     int ret = -1;
     int32_t op_errno = ENOMEM;
+    int8_t last_fsync = 0;
 
     AFR_ERROR_OUT_IF_FDCTX_INVALID(fd, this, op_errno, out);
     transaction_frame = copy_frame(frame);
@@ -2516,10 +2517,16 @@ afr_fsync(call_frame_t *frame, xlator_t *this, fd_t *fd, int32_t datasync,
     if (!local)
         goto out;
 
-    if (xdata)
+    if (xdata) {
         local->xdata_req = dict_copy_with_ref(xdata, NULL);
-    else
+        if (dict_get_int8(xdata, "last-fsync", &last_fsync) == 0) {
+            if (last_fsync) {
+                local->transaction.disable_delayed_post_op = _gf_true;
+            }
+        }
+    } else {
         local->xdata_req = dict_new();
+    }
 
     if (!local->xdata_req)
         goto out;