From e84346b726ea90a8ed470bc81c4136a7b8710ea5 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 11 May 2010 21:20:38 -0700 Subject: [PATCH] ceph: preserve seq # on requeued messages after transient transport errors If the tcp connection drops and we reconnect to reestablish a stateful session (with the mds), we need to resend previously sent (and possibly received) messages with the _same_ seq # so that they can be dropped on the other end if needed. Only assign a new seq once after the message is queued. Signed-off-by: Sage Weil --- fs/ceph/messenger.c | 11 ++++++++++- fs/ceph/messenger.h | 1 + 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/fs/ceph/messenger.c b/fs/ceph/messenger.c index a3a8f36..cd4fadb 100644 --- a/fs/ceph/messenger.c +++ b/fs/ceph/messenger.c @@ -492,7 +492,14 @@ static void prepare_write_message(struct ceph_connection *con) list_move_tail(&m->list_head, &con->out_sent); } - m->hdr.seq = cpu_to_le64(++con->out_seq); + /* + * only assign outgoing seq # if we haven't sent this message + * yet. if it is requeued, resend with it's original seq. + */ + if (m->needs_out_seq) { + m->hdr.seq = cpu_to_le64(++con->out_seq); + m->needs_out_seq = false; + } dout("prepare_write_message %p seq %lld type %d len %d+%d+%d %d pgs\n", m, con->out_seq, le16_to_cpu(m->hdr.type), @@ -1986,6 +1993,8 @@ void ceph_con_send(struct ceph_connection *con, struct ceph_msg *msg) BUG_ON(msg->front.iov_len != le32_to_cpu(msg->hdr.front_len)); + msg->needs_out_seq = true; + /* queue */ mutex_lock(&con->mutex); BUG_ON(!list_empty(&msg->list_head)); diff --git a/fs/ceph/messenger.h b/fs/ceph/messenger.h index a343dae..a5caf91 100644 --- a/fs/ceph/messenger.h +++ b/fs/ceph/messenger.h @@ -86,6 +86,7 @@ struct ceph_msg { struct kref kref; bool front_is_vmalloc; bool more_to_follow; + bool needs_out_seq; int front_max; struct ceph_msgpool *pool; -- 1.8.2.3