[Sheepdog] [PATCH] collie: setup cluster_info correctly when collie restarts

MORITA Kazutaka morita.kazutaka at lab.ntt.co.jp
Wed May 12 03:58:55 CEST 2010


Signed-off-by: MORITA Kazutaka <morita.kazutaka at lab.ntt.co.jp>
---
 collie/group.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/collie/group.c b/collie/group.c
index 349ce9a..a5479f9 100644
--- a/collie/group.c
+++ b/collie/group.c
@@ -1166,6 +1166,8 @@ static void __sd_confchg(struct cpg_event *cevent)
 		nr_entries = ARRAY_SIZE(entries);
 		ret = read_epoch(&epoch, &ctime, entries, &nr_entries);
 		if (ret == SD_RES_SUCCESS) {
+			sys->epoch = epoch;
+			msg.ctime = ctime;
 			get_cluster_status(&msg.header.from, entries, nr_entries,
 					   ctime, epoch, &msg.cluster_status);
 		} else
-- 
1.5.6.5




More information about the sheepdog mailing list