import ceph nautilus 14.2.2

[ceph.git] / ceph / src / mds / Server.cc

diff --git a/ceph/src/mds/Server.cc b/ceph/src/mds/Server.cc
index 8bfd19cbcfaad791a21feeb7348724b1ab4faeeb..b125937a4ff048dbc0ccf4edee7d2d955274fee1 100644 (file)
--- a/ceph/src/mds/Server.cc
+++ b/ceph/src/mds/Server.cc
@@ -193,6 +193,7 @@ Server::Server(MDSRank *m) :
   terminating_sessions(false),
   recall_throttle(g_conf().get_val<double>("mds_recall_max_decay_rate"))
 {
+  cap_revoke_eviction_timeout = g_conf().get_val<double>("mds_cap_revoke_eviction_timeout");
   supported_features = feature_bitset_t(CEPHFS_FEATURES_MDS_SUPPORTED);
 }
 
@@ -1108,9 +1109,14 @@ void Server::kill_session(Session *session, Context *on_safe)
 
 size_t Server::apply_blacklist(const std::set<entity_addr_t> &blacklist)
 {
-  std::list<Session*> victims;
+  bool prenautilus = mds->objecter->with_osdmap(
+      [&](const OSDMap& o) {
+       return o.require_osd_release < CEPH_RELEASE_NAUTILUS;
+      });
+
+  std::vector<Session*> victims;
   const auto& sessions = mds->sessionmap.get_sessions();
-  for (const auto& p : sessions)  {
+  for (const auto& p : sessions) {
     if (!p.first.is_client()) {
       // Do not apply OSDMap blacklist to MDS daemons, we find out
       // about their death via MDSMap.
@@ -1118,8 +1124,19 @@ size_t Server::apply_blacklist(const std::set<entity_addr_t> &blacklist)
     }
 
     Session *s = p.second;
-    if (blacklist.count(s->info.inst.addr)) {
+    auto inst_addr = s->info.inst.addr;
+    // blacklist entries are always TYPE_ANY for nautilus+
+    inst_addr.set_type(entity_addr_t::TYPE_ANY);
+    if (blacklist.count(inst_addr)) {
       victims.push_back(s);
+      continue;
+    }
+    if (prenautilus) {
+      // ...except pre-nautilus, they were TYPE_LEGACY
+      inst_addr.set_type(entity_addr_t::TYPE_LEGACY);
+      if (blacklist.count(inst_addr)) {
+       victims.push_back(s);
+      }
     }
   }
 
@@ -1199,7 +1216,8 @@ void Server::handle_client_reconnect(const MClientReconnect::const_ref &m)
          << (m->has_more() ? " (more)" : "") << dendl;
   client_t from = m->get_source().num();
   Session *session = mds->get_session(m);
-  ceph_assert(session);
+  if (!session)
+    return;
 
   if (!mds->is_reconnect() && mds->get_want_state() == CEPH_MDS_STATE_RECONNECT) {
     dout(10) << " we're almost in reconnect state (mdsmap delivery race?); waiting" << dendl;
@@ -1940,7 +1958,7 @@ void Server::reply_client_request(MDRequestRef& mdr, const MClientReply::ref &re
     mds->logger->inc(l_mds_reply);
     utime_t lat = ceph_clock_now() - mdr->client_request->get_recv_stamp();
     mds->logger->tinc(l_mds_reply_latency, lat);
-    if (client_inst.name.is_client()) {
+    if (session && client_inst.name.is_client()) {
       mds->sessionmap.hit_session(session);
     }
     perf_gather_op_latency(req, lat);
@@ -1956,7 +1974,7 @@ void Server::reply_client_request(MDRequestRef& mdr, const MClientReply::ref &re
   mdcache->request_drop_non_rdlocks(mdr);
 
   // reply at all?
-  if (!(client_inst.name.is_mds() || !session)) {
+  if (session && !client_inst.name.is_mds()) {
     // send reply.
     if (!did_early_reply &&   // don't issue leases if we sent an earlier reply already
        (tracei || tracedn)) {
@@ -3182,7 +3200,7 @@ void Server::apply_allocated_inos(MDRequestRef& mdr, Session *session)
     ceph_assert(session);
     session->pending_prealloc_inos.subtract(mdr->prealloc_inos);
     session->info.prealloc_inos.insert(mdr->prealloc_inos);
-    mds->sessionmap.mark_dirty(session);
+    mds->sessionmap.mark_dirty(session, !mdr->used_prealloc_ino);
     mds->inotable->apply_alloc_ids(mdr->prealloc_inos);
   }
   if (mdr->used_prealloc_ino) {
@@ -5830,6 +5848,11 @@ void Server::handle_client_mkdir(MDRequestRef& mdr)
   le->metablob.add_opened_ino(newi->ino());
 
   journal_and_reply(mdr, newi, dn, le, new C_MDS_mknod_finish(this, mdr, dn, newi));
+
+  // We hit_dir (via hit_inode) in our finish callback, but by then we might
+  // have overshot the split size (multiple mkdir in flight), so here is
+  // an early chance to split the dir if this mkdir makes it oversized.
+  mds->balancer->maybe_fragment(dir, false);
 }
 
 
@@ -7756,6 +7779,7 @@ void Server::handle_client_rename(MDRequestRef& mdr)
   C_MDS_rename_finish *fin = new C_MDS_rename_finish(this, mdr, srcdn, destdn, straydn);
 
   journal_and_reply(mdr, srci, destdn, le, fin);
+  mds->balancer->maybe_fragment(destdn->get_dir(), false);
 }