[Sheepdog] [PATCH 2/4] sheep: move eprintf in update_cluster_info()

[Liu Yuan]

From: Liu Yuan <tailai.ly at taobao.com>

If node is already joined, we lose the changce to see log information. So
move it forward in a better place.

Signed-off-by: Liu Yuan <tailai.ly at taobao.com>
---
 sheep/group.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/sheep/group.c b/sheep/group.c
index ee12cf3..032d44c 100644
--- a/sheep/group.c
+++ b/sheep/group.c
@@ -606,6 +606,7 @@ static void update_cluster_info(struct join_message *msg)
 	int ret, nr_nodes = msg->nr_nodes;
 	struct sheepdog_node_list_entry entry[SD_MAX_NODES];
 
+	eprintf("status = %d, epoch = %d, %d, %d\n", msg->cluster_status, msg->epoch, msg->result, sys->join_finished);
 	if (msg->result != SD_RES_SUCCESS) {
 		if (is_myself(msg->header.from.addr, msg->header.from.port)) {
 			eprintf("failed to join sheepdog, %d\n", msg->result);
@@ -639,7 +640,6 @@ static void update_cluster_info(struct join_message *msg)
 
 	sys->join_finished = 1;
 
-	eprintf("system status = %d, epoch = %d\n", msg->cluster_status, sys->epoch);
 	if (msg->cluster_status == SD_STATUS_OK && msg->inc_epoch) {
 		nr_nodes = get_ordered_sd_node_list(entry);
 
-- 
1.7.5.1