summaryrefslogtreecommitdiffstats
path: root/fs/ceph
diff options
context:
space:
mode:
authorGreg Farnum <gregf@hq.newdream.net>2010-06-22 15:58:01 -0700
committerSage Weil <sage@newdream.net>2010-08-01 20:11:41 -0700
commite55b71f802fd448a79275ba7b263fe1a8639be5f (patch)
treefa17d2b0fbf1bc5e76a5e9ddac6c19a6bbe8d2bc /fs/ceph
parent2bc50259fa0aa1868f8b2ba1d374406cb3c57f72 (diff)
downloadop-kernel-dev-e55b71f802fd448a79275ba7b263fe1a8639be5f.zip
op-kernel-dev-e55b71f802fd448a79275ba7b263fe1a8639be5f.tar.gz
ceph: handle ESTALE properly; on receipt send to authority if it wasn't
Signed-off-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
Diffstat (limited to 'fs/ceph')
-rw-r--r--fs/ceph/mds_client.c41
-rw-r--r--fs/ceph/mds_client.h2
2 files changed, 35 insertions, 8 deletions
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index a546e0d..34d215f 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -1628,6 +1628,15 @@ static int __prepare_send_request(struct ceph_mds_client *mdsc,
req->r_mds = mds;
req->r_attempts++;
+ if (req->r_inode) {
+ struct ceph_cap *cap =
+ ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);
+
+ if (cap)
+ req->r_sent_on_mseq = cap->mseq;
+ else
+ req->r_sent_on_mseq = -1;
+ }
dout("prepare_send_request %p tid %lld %s (attempt %d)\n", req,
req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);
@@ -1962,21 +1971,39 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
result = le32_to_cpu(head->result);
/*
- * Tolerate 2 consecutive ESTALEs from the same mds.
- * FIXME: we should be looking at the cap migrate_seq.
+ * Handle an ESTALE
+ * if we're not talking to the authority, send to them
+ * if the authority has changed while we weren't looking,
+ * send to new authority
+ * Otherwise we just have to return an ESTALE
*/
if (result == -ESTALE) {
- req->r_direct_mode = USE_AUTH_MDS;
- req->r_num_stale++;
- if (req->r_num_stale <= 2) {
+ dout("got ESTALE on request %llu", req->r_tid);
+ if (!req->r_inode) ; //do nothing; not an authority problem
+ else if (req->r_direct_mode != USE_AUTH_MDS) {
+ dout("not using auth, setting for that now");
+ req->r_direct_mode = USE_AUTH_MDS;
__do_request(mdsc, req);
mutex_unlock(&mdsc->mutex);
goto out;
+ } else {
+ struct ceph_inode_info *ci = ceph_inode(req->r_inode);
+ struct ceph_cap *cap =
+ ceph_get_cap_for_mds(ci, req->r_mds);;
+
+ dout("already using auth");
+ if ((!cap || cap != ci->i_auth_cap) ||
+ (cap->mseq != req->r_sent_on_mseq)) {
+ dout("but cap changed, so resending");
+ __do_request(mdsc, req);
+ mutex_unlock(&mdsc->mutex);
+ goto out;
+ }
}
- } else {
- req->r_num_stale = 0;
+ dout("have to return ESTALE on request %llu", req->r_tid);
}
+
if (head->safe) {
req->r_got_safe = true;
__unregister_request(mdsc, req);
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index c86be30..ab7e89f 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -208,8 +208,8 @@ struct ceph_mds_request {
int r_attempts; /* resend attempts */
int r_num_fwd; /* number of forward attempts */
- int r_num_stale;
int r_resend_mds; /* mds to resend to next, if any*/
+ u32 r_sent_on_mseq; /* cap mseq request was sent at*/
struct kref r_kref;
struct list_head r_wait;
OpenPOWER on IntegriCloud