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

// -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
// vim: ts=8 sw=2 smarttab
/*
 * Ceph - scalable distributed file system
 *
 * Copyright (C) 2015 Red Hat
 *
 * This is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License version 2.1, as published by the Free Software
 * Foundation.  See file COPYING.
 *
 */

#include <string_view>

#include "common/debug.h"
#include "common/errno.h"

#include "messages/MClientRequestForward.h"
#include "messages/MMDSLoadTargets.h"
#include "messages/MMDSTableRequest.h"

#include "mgr/MgrClient.h"

#include "MDSDaemon.h"
#include "MDSMap.h"
#include "SnapClient.h"
#include "SnapServer.h"
#include "MDBalancer.h"
#include "Migrator.h"
#include "Locker.h"
#include "InoTable.h"
#include "mon/MonClient.h"
#include "common/HeartbeatMap.h"
#include "ScrubStack.h"


#include "MDSRank.h"

#define dout_context g_ceph_context
#define dout_subsys ceph_subsys_mds
#undef dout_prefix
#define dout_prefix *_dout << "mds." << whoami << '.' << incarnation << ' '
using TOPNSPC::common::cmd_getval;
class C_Flush_Journal : public MDSInternalContext {
public:
  C_Flush_Journal(MDCache *mdcache, MDLog *mdlog, MDSRank *mds,
                  std::ostream *ss, Context *on_finish)
    : MDSInternalContext(mds),
      mdcache(mdcache), mdlog(mdlog), ss(ss), on_finish(on_finish),
      whoami(mds->whoami), incarnation(mds->incarnation) {
  }

  void send() {
    assert(ceph_mutex_is_locked(mds->mds_lock));

    dout(20) << __func__ << dendl;

    if (mdcache->is_readonly()) {
      dout(5) << __func__ << ": read-only FS" << dendl;
      complete(-EROFS);
      return;
    }

    if (!mds->is_active()) {
      dout(5) << __func__ << ": MDS not active, no-op" << dendl;
      complete(0);
      return;
    }

    flush_mdlog();
  }

private:

  void flush_mdlog() {
    dout(20) << __func__ << dendl;

    // I need to seal off the current segment, and then mark all
    // previous segments for expiry
    mdlog->start_new_segment();

    Context *ctx = new LambdaContext([this](int r) {
        handle_flush_mdlog(r);
      });

    // Flush initially so that all the segments older than our new one
    // will be elegible for expiry
    mdlog->flush();
    mdlog->wait_for_safe(new MDSInternalContextWrapper(mds, ctx));
  }

  void handle_flush_mdlog(int r) {
    dout(20) << __func__ << ": r=" << r << dendl;

    if (r != 0) {
      *ss << "Error " << r << " (" << cpp_strerror(r) << ") while flushing journal";
      complete(r);
      return;
    }

    clear_mdlog();
  }

  void clear_mdlog() {
    dout(20) << __func__ << dendl;

    Context *ctx = new LambdaContext([this](int r) {
        handle_clear_mdlog(r);
      });

    // Because we may not be the last wait_for_safe context on MDLog,
    // and subsequent contexts might wake up in the middle of our
    // later trim_all and interfere with expiry (by e.g. marking
    // dirs/dentries dirty on previous log segments), we run a second
    // wait_for_safe here. See #10368
    mdlog->wait_for_safe(new MDSInternalContextWrapper(mds, ctx));
  }

  void handle_clear_mdlog(int r) {
    dout(20) << __func__ << ": r=" << r << dendl;

    if (r != 0) {
      *ss << "Error " << r << " (" << cpp_strerror(r) << ") while flushing journal";
      complete(r);
      return;
    }

    trim_mdlog();
  }

  void trim_mdlog() {
    // Put all the old log segments into expiring or expired state
    dout(5) << __func__ << ": beginning segment expiry" << dendl;

    int ret = mdlog->trim_all();
    if (ret != 0) {
      *ss << "Error " << ret << " (" << cpp_strerror(ret) << ") while trimming log";
      complete(ret);
      return;
    }

    expire_segments();
  }

  void expire_segments() {
    dout(20) << __func__ << dendl;

    // Attach contexts to wait for all expiring segments to expire
    MDSGatherBuilder *expiry_gather = new MDSGatherBuilder(g_ceph_context);

    const auto &expiring_segments = mdlog->get_expiring_segments();
    for (auto p : expiring_segments) {
      p->wait_for_expiry(expiry_gather->new_sub());
    }
    dout(5) << __func__ << ": waiting for " << expiry_gather->num_subs_created()
            << " segments to expire" << dendl;

    if (!expiry_gather->has_subs()) {
      trim_segments();
      delete expiry_gather;
      return;
    }

    Context *ctx = new LambdaContext([this](int r) {
        handle_expire_segments(r);
      });
    expiry_gather->set_finisher(new MDSInternalContextWrapper(mds, ctx));
    expiry_gather->activate();
  }

  void handle_expire_segments(int r) {
    dout(20) << __func__ << ": r=" << r << dendl;

    ceph_assert(r == 0); // MDLog is not allowed to raise errors via
                         // wait_for_expiry
    trim_segments();
  }

  void trim_segments() {
    dout(20) << __func__ << dendl;

    Context *ctx = new C_OnFinisher(new LambdaContext([this](int) {
          std::lock_guard locker(mds->mds_lock);
          trim_expired_segments();
        }), mds->finisher);
    ctx->complete(0);
  }

  void trim_expired_segments() {
    dout(5) << __func__ << ": expiry complete, expire_pos/trim_pos is now "
            << std::hex << mdlog->get_journaler()->get_expire_pos() << "/"
            << mdlog->get_journaler()->get_trimmed_pos() << dendl;

    // Now everyone I'm interested in is expired
    mdlog->trim_expired_segments();

    dout(5) << __func__ << ": trim complete, expire_pos/trim_pos is now "
            << std::hex << mdlog->get_journaler()->get_expire_pos() << "/"
            << mdlog->get_journaler()->get_trimmed_pos() << dendl;

    write_journal_head();
  }

  void write_journal_head() {
    dout(20) << __func__ << dendl;

    Context *ctx = new LambdaContext([this](int r) {
        std::lock_guard locker(mds->mds_lock);
        handle_write_head(r);
      });
    // Flush the journal header so that readers will start from after
    // the flushed region
    mdlog->get_journaler()->write_head(ctx);
  }

  void handle_write_head(int r) {
    if (r != 0) {
      *ss << "Error " << r << " (" << cpp_strerror(r) << ") while writing header";
    } else {
      dout(5) << __func__ << ": write_head complete, all done!" << dendl;
    }

    complete(r);
  }

  void finish(int r) override {
    dout(20) << __func__ << ": r=" << r << dendl;
    on_finish->complete(r);
  }

  MDCache *mdcache;
  MDLog *mdlog;
  std::ostream *ss;
  Context *on_finish;

  // so as to use dout
  mds_rank_t whoami;
  int incarnation;
};

class C_Drop_Cache : public MDSInternalContext {
public:
  C_Drop_Cache(Server *server, MDCache *mdcache, MDLog *mdlog,
               MDSRank *mds, uint64_t recall_timeout,
               Formatter *f, Context *on_finish)
    : MDSInternalContext(mds),
      server(server), mdcache(mdcache), mdlog(mdlog),
      recall_timeout(recall_timeout), recall_start(mono_clock::now()),
      f(f), on_finish(on_finish),
      whoami(mds->whoami), incarnation(mds->incarnation) {
  }

  void send() {
    // not really a hard requirement here, but lets ensure this in
    // case we change the logic here.
    assert(ceph_mutex_is_locked(mds->mds_lock));

    dout(20) << __func__ << dendl;
    f->open_object_section("result");
    recall_client_state();
  }

private:
  // context which completes itself (with -ETIMEDOUT) after a specified
  // timeout or when explicitly completed, whichever comes first. Note
  // that the context does not detroy itself after completion -- it
  // needs to be explicitly freed.
  class C_ContextTimeout : public MDSInternalContext {
  public:
    C_ContextTimeout(MDSRank *mds, uint64_t timeout, Context *on_finish)
      : MDSInternalContext(mds),
        timeout(timeout),
        on_finish(on_finish) {
    }
    ~C_ContextTimeout() {
      ceph_assert(timer_task == nullptr);
    }

    void start_timer() {
      if (!timeout) {
        return;
      }

      timer_task = new LambdaContext([this](int) {
          timer_task = nullptr;
          complete(-ETIMEDOUT);
        });
      mds->timer.add_event_after(timeout, timer_task);
    }

    void finish(int r) override {
      Context *ctx = nullptr;
      {
        std::lock_guard locker(lock);
        std::swap(on_finish, ctx);
      }
      if (ctx != nullptr) {
        ctx->complete(r);
      }
    }
    void complete(int r) override {
      if (timer_task != nullptr) {
        mds->timer.cancel_event(timer_task);
      }

      finish(r);
    }

    uint64_t timeout;
    ceph::mutex lock = ceph::make_mutex("mds::context::timeout");
    Context *on_finish = nullptr;
    Context *timer_task = nullptr;
  };

  auto do_trim() {
    auto [throttled, count] = mdcache->trim(UINT64_MAX);
    dout(10) << __func__
             << (throttled ? " (throttled)" : "")
             << " trimmed " << count << " caps" << dendl;
    dentries_trimmed += count;
    return std::make_pair(throttled, count);
  }

  void recall_client_state() {
    dout(20) << __func__ << dendl;
    auto now = mono_clock::now();
    auto duration = std::chrono::duration<double>(now-recall_start).count();

    MDSGatherBuilder *gather = new MDSGatherBuilder(g_ceph_context);
    auto flags = Server::RecallFlags::STEADY|Server::RecallFlags::TRIM;
    auto [throttled, count] = server->recall_client_state(gather, flags);
    dout(10) << __func__
             << (throttled ? " (throttled)" : "")
             << " recalled " << count << " caps" << dendl;

    caps_recalled += count;
    if ((throttled || count > 0) && (recall_timeout == 0 || duration < recall_timeout)) {
      C_ContextTimeout *ctx = new C_ContextTimeout(
        mds, 1, new LambdaContext([this](int r) {
          recall_client_state();
      }));
      ctx->start_timer();
      gather->set_finisher(new MDSInternalContextWrapper(mds, ctx));
      gather->activate();
      mdlog->flush(); /* use down-time to incrementally flush log */
      do_trim(); /* use down-time to incrementally trim cache */
    } else {
      if (!gather->has_subs()) {
        delete gather;
        return handle_recall_client_state(0);
      } else if (recall_timeout > 0 && duration > recall_timeout) {
        gather->set_finisher(new C_MDSInternalNoop);
        gather->activate();
        return handle_recall_client_state(-ETIMEDOUT);
      } else {
        uint64_t remaining = (recall_timeout == 0 ? 0 : recall_timeout-duration);
        C_ContextTimeout *ctx = new C_ContextTimeout(
          mds, remaining, new LambdaContext([this](int r) {
              handle_recall_client_state(r);
            }));

        ctx->start_timer();
        gather->set_finisher(new MDSInternalContextWrapper(mds, ctx));
        gather->activate();
      }
    }
  }

  void handle_recall_client_state(int r) {
    dout(20) << __func__ << ": r=" << r << dendl;

    // client recall section
    f->open_object_section("client_recall");
    f->dump_int("return_code", r);
    f->dump_string("message", cpp_strerror(r));
    f->dump_int("recalled", caps_recalled);
    f->close_section();

    // we can still continue after recall timeout
    flush_journal();
  }

  void flush_journal() {
    dout(20) << __func__ << dendl;

    Context *ctx = new LambdaContext([this](int r) {
        handle_flush_journal(r);
      });

    C_Flush_Journal *flush_journal = new C_Flush_Journal(mdcache, mdlog, mds, &ss, ctx);
    flush_journal->send();
  }

  void handle_flush_journal(int r) {
    dout(20) << __func__ << ": r=" << r << dendl;

    if (r != 0) {
      cmd_err(f, ss.str());
      complete(r);
      return;
    }

    // journal flush section
    f->open_object_section("flush_journal");
    f->dump_int("return_code", r);
    f->dump_string("message", ss.str());
    f->close_section();

    trim_cache();
  }

  void trim_cache() {
    dout(20) << __func__ << dendl;

    auto [throttled, count] = do_trim();
    if (throttled && count > 0) {
      auto timer = new LambdaContext([this](int) {
        trim_cache();
      });
      mds->timer.add_event_after(1.0, timer);
    } else {
      cache_status();
    }
  }

  void cache_status() {
    dout(20) << __func__ << dendl;

    f->open_object_section("trim_cache");
    f->dump_int("trimmed", dentries_trimmed);
    f->close_section();

    // cache status section
    mdcache->cache_status(f);

    complete(0);
  }

  void finish(int r) override {
    dout(20) << __func__ << ": r=" << r << dendl;

    auto d = std::chrono::duration<double>(mono_clock::now()-recall_start);
    f->dump_float("duration", d.count());

    f->close_section();
    on_finish->complete(r);
  }

  Server *server;
  MDCache *mdcache;
  MDLog *mdlog;
  uint64_t recall_timeout;
  mono_time recall_start;
  Formatter *f;
  Context *on_finish;

  int retval = 0;
  std::stringstream ss;
  uint64_t caps_recalled = 0;
  uint64_t dentries_trimmed = 0;

  // so as to use dout
  mds_rank_t whoami;
  int incarnation;

  void cmd_err(Formatter *f, std::string_view err) {
    f->reset();
    f->open_object_section("result");
    f->dump_string("error", err);
    f->close_section();
  }
};

MDSRank::MDSRank(
    mds_rank_t whoami_,
    ceph::mutex &mds_lock_,
    LogChannelRef &clog_,
    SafeTimer &timer_,
    Beacon &beacon_,
    std::unique_ptr<MDSMap>& mdsmap_,
    Messenger *msgr,
    MonClient *monc_,
    MgrClient *mgrc,
    Context *respawn_hook_,
    Context *suicide_hook_) :
    cct(msgr->cct), mds_lock(mds_lock_), clog(clog_),
    timer(timer_), mdsmap(mdsmap_),
    objecter(new Objecter(g_ceph_context, msgr, monc_, nullptr, 0, 0)),
    damage_table(whoami_), sessionmap(this),
    op_tracker(g_ceph_context, g_conf()->mds_enable_op_tracker,
               g_conf()->osd_num_op_tracker_shard),
    progress_thread(this), whoami(whoami_),
    purge_queue(g_ceph_context, whoami_,
      mdsmap_->get_metadata_pool(), objecter,
      new LambdaContext([this](int r) {
          std::lock_guard l(mds_lock);
          handle_write_error(r);
        }
      )
    ),
    beacon(beacon_),
    messenger(msgr), monc(monc_), mgrc(mgrc),
    respawn_hook(respawn_hook_),
    suicide_hook(suicide_hook_),
    starttime(mono_clock::now())
{
  hb = g_ceph_context->get_heartbeat_map()->add_worker("MDSRank", pthread_self());

  purge_queue.update_op_limit(*mdsmap);

  objecter->unset_honor_pool_full();

  finisher = new Finisher(cct, "MDSRank", "MR_Finisher");

  mdcache = new MDCache(this, purge_queue);
  mdlog = new MDLog(this);
  balancer = new MDBalancer(this, messenger, monc);

  scrubstack = new ScrubStack(mdcache, clog, finisher);

  inotable = new InoTable(this);
  snapserver = new SnapServer(this, monc);
  snapclient = new SnapClient(this);

  server = new Server(this);
  locker = new Locker(this, mdcache);

  op_tracker.set_complaint_and_threshold(cct->_conf->mds_op_complaint_time,
                                         cct->_conf->mds_op_log_threshold);
  op_tracker.set_history_size_and_duration(cct->_conf->mds_op_history_size,
                                           cct->_conf->mds_op_history_duration);

  schedule_update_timer_task();
}

MDSRank::~MDSRank()
{
  if (hb) {
    g_ceph_context->get_heartbeat_map()->remove_worker(hb);
  }

  if (scrubstack) { delete scrubstack; scrubstack = NULL; }
  if (mdcache) { delete mdcache; mdcache = NULL; }
  if (mdlog) { delete mdlog; mdlog = NULL; }
  if (balancer) { delete balancer; balancer = NULL; }
  if (inotable) { delete inotable; inotable = NULL; }
  if (snapserver) { delete snapserver; snapserver = NULL; }
  if (snapclient) { delete snapclient; snapclient = NULL; }

  if (server) { delete server; server = 0; }
  if (locker) { delete locker; locker = 0; }

  if (logger) {
    g_ceph_context->get_perfcounters_collection()->remove(logger);
    delete logger;
    logger = 0;
  }
  if (mlogger) {
    g_ceph_context->get_perfcounters_collection()->remove(mlogger);
    delete mlogger;
    mlogger = 0;
  }

  delete finisher;
  finisher = NULL;

  delete suicide_hook;
  suicide_hook = NULL;

  delete respawn_hook;
  respawn_hook = NULL;

  delete objecter;
  objecter = nullptr;
}

void MDSRankDispatcher::init()
{
  objecter->init();
  messenger->add_dispatcher_head(objecter);

  objecter->start();

  update_log_config();
  create_logger();

  // Expose the OSDMap (already populated during MDS::init) to anyone
  // who is interested in it.
  handle_osd_map();

  progress_thread.create("mds_rank_progr");

  purge_queue.init();

  finisher->start();
}

void MDSRank::update_targets()
{
  // get MonMap's idea of my export_targets
  const set<mds_rank_t>& map_targets = mdsmap->get_mds_info(get_nodeid()).export_targets;

  dout(20) << "updating export targets, currently " << map_targets.size() << " ranks are targets" << dendl;

  bool send = false;
  set<mds_rank_t> new_map_targets;

  auto it = export_targets.begin();
  while (it != export_targets.end()) {
    mds_rank_t rank = it->first;
    auto &counter = it->second;
    dout(20) << "export target mds." << rank << " is " << counter << dendl;

    double val = counter.get();
    if (val <= 0.01) {
      dout(15) << "export target mds." << rank << " is no longer an export target" << dendl;
      export_targets.erase(it++);
      send = true;
      continue;
    }
    if (!map_targets.count(rank)) {
      dout(15) << "export target mds." << rank << " not in map's export_targets" << dendl;
      send = true;
    }
    new_map_targets.insert(rank);
    it++;
  }
  if (new_map_targets.size() < map_targets.size()) {
    dout(15) << "export target map holds stale targets, sending update" << dendl;
    send = true;
  }

  if (send) {
    dout(15) << "updating export_targets, now " << new_map_targets.size() << " ranks are targets" << dendl;
    auto m = make_message<MMDSLoadTargets>(mds_gid_t(monc->get_global_id()), new_map_targets);
    monc->send_mon_message(m.detach());
  }
}

void MDSRank::hit_export_target(mds_rank_t rank, double amount)
{
  double rate = g_conf()->mds_bal_target_decay;
  if (amount < 0.0) {
    amount = 100.0/g_conf()->mds_bal_target_decay; /* a good default for "i am trying to keep this export_target active" */
  }
  auto em = export_targets.emplace(std::piecewise_construct, std::forward_as_tuple(rank), std::forward_as_tuple(DecayRate(rate)));
  auto &counter = em.first->second;
  counter.hit(amount);
  if (em.second) {
    dout(15) << "hit export target (new) is " << counter << dendl;
  } else {
    dout(15) << "hit export target is " << counter << dendl;
  }
}

class C_MDS_MonCommand : public MDSInternalContext {
  std::string cmd;
public:
  std::string outs;
  C_MDS_MonCommand(MDSRank *m, std::string_view c)
    : MDSInternalContext(m), cmd(c) {}
  void finish(int r) override {
    mds->_mon_command_finish(r, cmd, outs);
  }
};

void MDSRank::_mon_command_finish(int r, std::string_view cmd, std::string_view outs)
{
  if (r < 0) {
    dout(0) << __func__ << ": mon command " << cmd << " failed with errno " << r
            << " (" << outs << ")" << dendl;
  } else {
    dout(1) << __func__ << ": mon command " << cmd << " succeed" << dendl;
  }
}

void MDSRank::set_mdsmap_multimds_snaps_allowed()
{
  static bool already_sent = false;
  if (already_sent)
    return;

  stringstream ss;
  ss << "{\"prefix\":\"fs set\", \"fs_name\":\"" <<  mdsmap->get_fs_name() << "\", ";
  ss << "\"var\":\"allow_multimds_snaps\", \"val\":\"true\", ";
  ss << "\"confirm\":\"--yes-i-am-really-a-mds\"}";
  std::vector<std::string> cmd = {ss.str()};

  dout(0) << __func__ << ": sending mon command: " << cmd[0] << dendl;

  C_MDS_MonCommand *fin = new C_MDS_MonCommand(this, cmd[0]);
  monc->start_mon_command(cmd, {}, nullptr, &fin->outs, new C_IO_Wrapper(this, fin));

  already_sent = true;
}

void MDSRank::mark_base_recursively_scrubbed(inodeno_t ino)
{
  if (mdsmap->get_tableserver() == whoami)
    snapserver->mark_base_recursively_scrubbed(ino);
}

void MDSRankDispatcher::tick()
{
  heartbeat_reset();

  if (beacon.is_laggy()) {
    dout(1) << "skipping upkeep work because connection to Monitors appears laggy" << dendl;
    return;
  }

  check_ops_in_flight();

  // Wake up thread in case we use to be laggy and have waiting_for_nolaggy
  // messages to progress.
  progress_thread.signal();

  // make sure mds log flushes, trims periodically
  mdlog->flush();

  // update average session uptime
  sessionmap.update_average_session_age();

  if (is_active() || is_stopping()) {
    mdlog->trim();  // NOT during recovery!
  }

  // ...
  if (is_cache_trimmable()) {
    server->find_idle_sessions();
    server->evict_cap_revoke_non_responders();
    locker->tick();
  }

  // log
  if (logger) {
    logger->set(l_mds_subtrees, mdcache->num_subtrees());
    mdcache->log_stat();
  }

  if (is_reconnect())
    server->reconnect_tick();

  if (is_active()) {
    balancer->tick();
    mdcache->find_stale_fragment_freeze();
    mdcache->migrator->find_stale_export_freeze();

    if (mdsmap->get_tableserver() == whoami) {
      snapserver->check_osd_map(false);
      // Filesystem was created by pre-mimic mds. Allow multi-active mds after
      // all old snapshots are deleted.
      if (!mdsmap->allows_multimds_snaps() &&
          snapserver->can_allow_multimds_snaps()) {
        set_mdsmap_multimds_snaps_allowed();
      }
    }
  }

  if (is_active() || is_stopping()) {
    update_targets();
  }

  // shut down?
  if (is_stopping()) {
    mdlog->trim();
    if (mdcache->shutdown_pass()) {
      uint64_t pq_progress = 0 ;
      uint64_t pq_total = 0;
      size_t pq_in_flight = 0;
      if (!purge_queue.drain(&pq_progress, &pq_total, &pq_in_flight)) {
        dout(7) << "shutdown_pass=true, but still waiting for purge queue"
                << dendl;
        // This takes unbounded time, so we must indicate progress
        // to the administrator: we do it in a slightly imperfect way
        // by sending periodic (tick frequency) clog messages while
        // in this state.
        clog->info() << "MDS rank " << whoami << " waiting for purge queue ("
          << std::dec << pq_progress << "/" << pq_total << " " << pq_in_flight
          << " files purging" << ")";
      } else {
        dout(7) << "shutdown_pass=true, finished w/ shutdown, moving to "
                   "down:stopped" << dendl;
        stopping_done();
      }
    }
    else {
      dout(7) << "shutdown_pass=false" << dendl;
    }
  }

  // Expose ourselves to Beacon to update health indicators
  beacon.notify_health(this);
}

void MDSRankDispatcher::shutdown()
{
  // It should never be possible for shutdown to get called twice, because
  // anyone picking up mds_lock checks if stopping is true and drops
  // out if it is.
  ceph_assert(stopping == false);
  stopping = true;

  dout(1) << __func__ << ": shutting down rank " << whoami << dendl;

  g_conf().remove_observer(this);

  timer.shutdown();

  // MDLog has to shut down before the finisher, because some of its
  // threads block on IOs that require finisher to complete.
  mdlog->shutdown();

  // shut down cache
  mdcache->shutdown();

  purge_queue.shutdown();

  mds_lock.unlock();
  finisher->stop(); // no flushing
  mds_lock.lock();

  if (objecter->initialized)
    objecter->shutdown();

  monc->shutdown();

  op_tracker.on_shutdown();

  progress_thread.shutdown();

  // release mds_lock for finisher/messenger threads (e.g.
  // MDSDaemon::ms_handle_reset called from Messenger).
  mds_lock.unlock();

  // shut down messenger
  messenger->shutdown();

  mds_lock.lock();

  // Workaround unclean shutdown: HeartbeatMap will assert if
  // worker is not removed (as we do in ~MDS), but ~MDS is not
  // always called after suicide.
  if (hb) {
    g_ceph_context->get_heartbeat_map()->remove_worker(hb);
    hb = NULL;
  }
}

/**
 * Helper for simple callbacks that call a void fn with no args.
 */
class C_MDS_VoidFn : public MDSInternalContext
{
  typedef void (MDSRank::*fn_ptr)();
  protected:
   fn_ptr fn;
  public:
  C_MDS_VoidFn(MDSRank *mds_, fn_ptr fn_)
    : MDSInternalContext(mds_), fn(fn_)
  {
    ceph_assert(mds_);
    ceph_assert(fn_);
  }

  void finish(int r) override
  {
    (mds->*fn)();
  }
};

int64_t MDSRank::get_metadata_pool()
{
    return mdsmap->get_metadata_pool();
}

MDSTableClient *MDSRank::get_table_client(int t)
{
  switch (t) {
  case TABLE_ANCHOR: return NULL;
  case TABLE_SNAP: return snapclient;
  default: ceph_abort();
  }
}

MDSTableServer *MDSRank::get_table_server(int t)
{
  switch (t) {
  case TABLE_ANCHOR: return NULL;
  case TABLE_SNAP: return snapserver;
  default: ceph_abort();
  }
}

void MDSRank::suicide()
{
  if (suicide_hook) {
    suicide_hook->complete(0);
    suicide_hook = NULL;
  }
}

void MDSRank::respawn()
{
  if (respawn_hook) {
    respawn_hook->complete(0);
    respawn_hook = NULL;
  }
}

void MDSRank::damaged()
{
  ceph_assert(whoami != MDS_RANK_NONE);
  ceph_assert(ceph_mutex_is_locked_by_me(mds_lock));

  beacon.set_want_state(*mdsmap, MDSMap::STATE_DAMAGED);
  monc->flush_log();  // Flush any clog error from before we were called
  beacon.notify_health(this);  // Include latest status in our swan song
  beacon.send_and_wait(g_conf()->mds_mon_shutdown_timeout);

  // It's okay if we timed out and the mon didn't get our beacon, because
  // another daemon (or ourselves after respawn) will eventually take the
  // rank and report DAMAGED again when it hits same problem we did.

  respawn();  // Respawn into standby in case mon has other work for us
}

void MDSRank::damaged_unlocked()
{
  std::lock_guard l(mds_lock);
  damaged();
}

void MDSRank::handle_write_error(int err)
{
  if (err == -EBLACKLISTED) {
    derr << "we have been blacklisted (fenced), respawning..." << dendl;
    respawn();
    return;
  }

  if (g_conf()->mds_action_on_write_error >= 2) {
    derr << "unhandled write error " << cpp_strerror(err) << ", suicide..." << dendl;
    respawn();
  } else if (g_conf()->mds_action_on_write_error == 1) {
    derr << "unhandled write error " << cpp_strerror(err) << ", force readonly..." << dendl;
    mdcache->force_readonly();
  } else {
    // ignore;
    derr << "unhandled write error " << cpp_strerror(err) << ", ignore..." << dendl;
  }
}

void *MDSRank::ProgressThread::entry()
{
  std::unique_lock l(mds->mds_lock);
  while (true) {
    cond.wait(l, [this] {
      return (mds->stopping ||
              !mds->finished_queue.empty() ||
              (!mds->waiting_for_nolaggy.empty() && !mds->beacon.is_laggy()));
    });

    if (mds->stopping) {
      break;
    }

    mds->_advance_queues();
  }

  return NULL;
}


void MDSRank::ProgressThread::shutdown()
{
  ceph_assert(ceph_mutex_is_locked_by_me(mds->mds_lock));
  ceph_assert(mds->stopping);

  if (am_self()) {
    // Stopping is set, we will fall out of our main loop naturally
  } else {
    // Kick the thread to notice mds->stopping, and join it
    cond.notify_all();
    mds->mds_lock.unlock();
    if (is_started())
      join();
    mds->mds_lock.lock();
  }
}

bool MDSRankDispatcher::ms_dispatch(const cref_t<Message> &m)
{
  if (m->get_source().is_client()) {
    Session *session = static_cast<Session*>(m->get_connection()->get_priv().get());
    if (session)
      session->last_seen = Session::clock::now();
  }

  inc_dispatch_depth();
  bool ret = _dispatch(m, true);
  dec_dispatch_depth();
  return ret;
}

bool MDSRank::_dispatch(const cref_t<Message> &m, bool new_msg)
{
  if (is_stale_message(m)) {
    return true;
  }
  // do not proceed if this message cannot be handled
  if (!is_valid_message(m)) {
    return false;
  }

  if (beacon.is_laggy()) {
    dout(5) << " laggy, deferring " << *m << dendl;
    waiting_for_nolaggy.push_back(m);
  } else if (new_msg && !waiting_for_nolaggy.empty()) {
    dout(5) << " there are deferred messages, deferring " << *m << dendl;
    waiting_for_nolaggy.push_back(m);
  } else {
    handle_message(m);
    heartbeat_reset();
  }

  if (dispatch_depth > 1)
    return true;

  // finish any triggered contexts
  _advance_queues();

  if (beacon.is_laggy()) {
    // We've gone laggy during dispatch, don't do any
    // more housekeeping
    return true;
  }

  // hack: thrash exports
  static utime_t start;
  utime_t now = ceph_clock_now();
  if (start == utime_t())
    start = now;
  /*double el = now - start;
  if (el > 30.0 &&
    el < 60.0)*/
  for (int i=0; i<g_conf()->mds_thrash_exports; i++) {
    set<mds_rank_t> s;
    if (!is_active()) break;
    mdsmap->get_mds_set(s, MDSMap::STATE_ACTIVE);
    if (s.size() < 2 || CInode::count() < 10)
      break;  // need peers for this to work.
    if (mdcache->migrator->get_num_exporting() > g_conf()->mds_thrash_exports * 5 ||
        mdcache->migrator->get_export_queue_size() > g_conf()->mds_thrash_exports * 10)
      break;

    dout(7) << "mds thrashing exports pass " << (i+1) << "/" << g_conf()->mds_thrash_exports << dendl;

    // pick a random dir inode
    CInode *in = mdcache->hack_pick_random_inode();

    auto&& ls = in->get_dirfrags();
    if (!ls.empty()) {  // must be an open dir.
      const auto& dir = ls[rand() % ls.size()];
      if (!dir->get_parent_dir()) continue;    // must be linked.
      if (!dir->is_auth()) continue;           // must be auth.

      mds_rank_t dest;
      do {
        int k = rand() % s.size();
        set<mds_rank_t>::iterator p = s.begin();
        while (k--) ++p;
        dest = *p;
      } while (dest == whoami);
      mdcache->migrator->export_dir_nicely(dir,dest);
    }
  }
  // hack: thrash fragments
  for (int i=0; i<g_conf()->mds_thrash_fragments; i++) {
    if (!is_active()) break;
    if (mdcache->get_num_fragmenting_dirs() > 5 * g_conf()->mds_thrash_fragments) break;
    dout(7) << "mds thrashing fragments pass " << (i+1) << "/" << g_conf()->mds_thrash_fragments << dendl;

    // pick a random dir inode
    CInode *in = mdcache->hack_pick_random_inode();

    auto&& ls = in->get_dirfrags();
    if (ls.empty()) continue;                // must be an open dir.
    CDir *dir = ls.front();
    if (!dir->get_parent_dir()) continue;    // must be linked.
    if (!dir->is_auth()) continue;           // must be auth.
    frag_t fg = dir->get_frag();
    if ((fg == frag_t() || (rand() % (1 << fg.bits()) == 0))) {
      mdcache->split_dir(dir, 1);
    } else {
      balancer->queue_merge(dir);
    }
  }

  // hack: force hash root?
  /*
  if (false &&
      mdcache->get_root() &&
      mdcache->get_root()->dir &&
      !(mdcache->get_root()->dir->is_hashed() ||
        mdcache->get_root()->dir->is_hashing())) {
    dout(0) << "hashing root" << dendl;
    mdcache->migrator->hash_dir(mdcache->get_root()->dir);
  }
  */

  update_mlogger();
  return true;
}

void MDSRank::update_mlogger()
{
  if (mlogger) {
    mlogger->set(l_mdm_ino, CInode::count());
    mlogger->set(l_mdm_dir, CDir::count());
    mlogger->set(l_mdm_dn, CDentry::count());
    mlogger->set(l_mdm_cap, Capability::count());
    mlogger->set(l_mdm_inoa, CInode::increments());
    mlogger->set(l_mdm_inos, CInode::decrements());
    mlogger->set(l_mdm_dira, CDir::increments());
    mlogger->set(l_mdm_dirs, CDir::decrements());
    mlogger->set(l_mdm_dna, CDentry::increments());
    mlogger->set(l_mdm_dns, CDentry::decrements());
    mlogger->set(l_mdm_capa, Capability::increments());
    mlogger->set(l_mdm_caps, Capability::decrements());
  }
}

// message types that the mds can handle
bool MDSRank::is_valid_message(const cref_t<Message> &m) {
  int port = m->get_type() & 0xff00;
  int type = m->get_type();

  if (port == MDS_PORT_CACHE ||
      port == MDS_PORT_MIGRATOR ||
      type == CEPH_MSG_CLIENT_SESSION ||
      type == CEPH_MSG_CLIENT_RECONNECT ||
      type == CEPH_MSG_CLIENT_RECLAIM ||
      type == CEPH_MSG_CLIENT_REQUEST ||
      type == MSG_MDS_SLAVE_REQUEST ||
      type == MSG_MDS_HEARTBEAT ||
      type == MSG_MDS_TABLE_REQUEST ||
      type == MSG_MDS_LOCK ||
      type == MSG_MDS_INODEFILECAPS ||
      type == CEPH_MSG_CLIENT_CAPS ||
      type == CEPH_MSG_CLIENT_CAPRELEASE ||
      type == CEPH_MSG_CLIENT_LEASE) {
    return true;
  }

  return false;
}

/*
 * lower priority messages we defer if we seem laggy
 */

#define ALLOW_MESSAGES_FROM(peers)                                      \
  do {                                                                  \
    if (m->get_connection() && (m->get_connection()->get_peer_type() & (peers)) == 0) { \
      dout(0) << __FILE__ << "." << __LINE__ << ": filtered out request, peer=" << m->get_connection()->get_peer_type() \
              << " allowing=" << #peers << " message=" << *m << dendl;  \
      return;                                                           \
    }                                                                   \
  } while (0)

void MDSRank::handle_message(const cref_t<Message> &m)
{
  int port = m->get_type() & 0xff00;

  switch (port) {
  case MDS_PORT_CACHE:
    ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_MDS);
    mdcache->dispatch(m);
    break;

  case MDS_PORT_MIGRATOR:
    ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_MDS);
    mdcache->migrator->dispatch(m);
    break;

  default:
    switch (m->get_type()) {
      // SERVER
    case CEPH_MSG_CLIENT_SESSION:
    case CEPH_MSG_CLIENT_RECONNECT:
    case CEPH_MSG_CLIENT_RECLAIM:
      ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_CLIENT);
      // fall-thru
    case CEPH_MSG_CLIENT_REQUEST:
      server->dispatch(m);
      break;
    case MSG_MDS_SLAVE_REQUEST:
      ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_MDS);
      server->dispatch(m);
      break;

    case MSG_MDS_HEARTBEAT:
      ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_MDS);
      balancer->proc_message(m);
      break;

    case MSG_MDS_TABLE_REQUEST:
      ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_MDS);
      {
        const cref_t<MMDSTableRequest> &req = ref_cast<MMDSTableRequest>(m);
        if (req->op < 0) {
          MDSTableClient *client = get_table_client(req->table);
          client->handle_request(req);
        } else {
           MDSTableServer *server = get_table_server(req->table);
           server->handle_request(req);
        }
      }
      break;

    case MSG_MDS_LOCK:
    case MSG_MDS_INODEFILECAPS:
      ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_MDS);
      locker->dispatch(m);
      break;

    case CEPH_MSG_CLIENT_CAPS:
    case CEPH_MSG_CLIENT_CAPRELEASE:
    case CEPH_MSG_CLIENT_LEASE:
      ALLOW_MESSAGES_FROM(CEPH_ENTITY_TYPE_CLIENT);
      locker->dispatch(m);
      break;

    default:
      derr << "unrecognized message " << *m << dendl;
    }
  }
}

/**
 * Advance finished_queue and waiting_for_nolaggy.
 *
 * Usually drain both queues, but may not drain waiting_for_nolaggy
 * if beacon is currently laggy.
 */
void MDSRank::_advance_queues()
{
  ceph_assert(ceph_mutex_is_locked_by_me(mds_lock));

  if (!finished_queue.empty()) {
    dout(7) << "mds has " << finished_queue.size() << " queued contexts" << dendl;
    while (!finished_queue.empty()) {
      auto fin = finished_queue.front();
      finished_queue.pop_front();

      dout(10) << " finish " << fin << dendl;
      fin->complete(0);

      heartbeat_reset();
    }
  }

  while (!waiting_for_nolaggy.empty()) {
    // stop if we're laggy now!
    if (beacon.is_laggy())
      break;

    cref_t<Message> old = waiting_for_nolaggy.front();
    waiting_for_nolaggy.pop_front();

    if (!is_stale_message(old)) {
      dout(7) << " processing laggy deferred " << *old << dendl;
      ceph_assert(is_valid_message(old));
      handle_message(old);
    }

    heartbeat_reset();
  }
}

/**
 * Call this when you take mds_lock, or periodically if you're going to
 * hold the lock for a long time (e.g. iterating over clients/inodes)
 */
void MDSRank::heartbeat_reset()
{
  // Any thread might jump into mds_lock and call us immediately
  // after a call to suicide() completes, in which case MDSRank::hb
  // has been freed and we are a no-op.
  if (!hb) {
      ceph_assert(stopping);
      return;
  }

  // NB not enabling suicide grace, because the mon takes care of killing us
  // (by blacklisting us) when we fail to send beacons, and it's simpler to
  // only have one way of dying.
  auto grace = g_conf().get_val<double>("mds_heartbeat_grace");
  g_ceph_context->get_heartbeat_map()->reset_timeout(hb, grace, 0);
}

bool MDSRank::is_stale_message(const cref_t<Message> &m) const
{
  // from bad mds?
  if (m->get_source().is_mds()) {
    mds_rank_t from = mds_rank_t(m->get_source().num());
    bool bad = false;
    if (mdsmap->is_down(from)) {
      bad = true;
    } else {
      // FIXME: this is a convoluted check.  we should be maintaining a nice
      // clean map of current ConnectionRefs for current mdses!!!
      auto c = messenger->connect_to(CEPH_ENTITY_TYPE_MDS,
                                     mdsmap->get_addrs(from));
      if (c != m->get_connection()) {
        bad = true;
        dout(5) << " mds." << from << " should be " << c << " "
                << c->get_peer_addrs() << " but this message is "
                << m->get_connection() << " " << m->get_source_addrs()
                << dendl;
      }
    }
    if (bad) {
      // bogus mds?
      if (m->get_type() == CEPH_MSG_MDS_MAP) {
        dout(5) << "got " << *m << " from old/bad/imposter mds " << m->get_source()
                << ", but it's an mdsmap, looking at it" << dendl;
      } else if (m->get_type() == MSG_MDS_CACHEEXPIRE &&
                 mdsmap->get_addrs(from) == m->get_source_addrs()) {
        dout(5) << "got " << *m << " from down mds " << m->get_source()
                << ", but it's a cache_expire, looking at it" << dendl;
      } else {
        dout(5) << "got " << *m << " from down/old/bad/imposter mds " << m->get_source()
                << ", dropping" << dendl;
        return true;
      }
    }
  }
  return false;
}

Session *MDSRank::get_session(const cref_t<Message> &m)
{
  // do not carry ref
  auto session = static_cast<Session *>(m->get_connection()->get_priv().get());
  if (session) {
    dout(20) << "get_session have " << session << " " << session->info.inst
             << " state " << session->get_state_name() << dendl;
    // Check if we've imported an open session since (new sessions start closed)
    if (session->is_closed()) {
      Session *imported_session = sessionmap.get_session(session->info.inst.name);
      if (imported_session && imported_session != session) {
        dout(10) << __func__ << " replacing connection bootstrap session "
                 << session << " with imported session " << imported_session
                 << dendl;
        imported_session->info.auth_name = session->info.auth_name;
        //assert(session->info.auth_name == imported_session->info.auth_name);
        ceph_assert(session->info.inst == imported_session->info.inst);
        imported_session->set_connection(session->get_connection().get());
        // send out any queued messages
        while (!session->preopen_out_queue.empty()) {
          imported_session->get_connection()->send_message2(std::move(session->preopen_out_queue.front()));
          session->preopen_out_queue.pop_front();
        }
        imported_session->auth_caps = session->auth_caps;
        imported_session->last_seen = session->last_seen;
        ceph_assert(session->get_nref() == 1);
        imported_session->get_connection()->set_priv(imported_session->get());
        session = imported_session;
      }
    }
  } else {
    dout(20) << "get_session dne for " << m->get_source_inst() << dendl;
  }
  return session;
}

void MDSRank::send_message(const ref_t<Message>& m, const ConnectionRef& c)
{
  ceph_assert(c);
  c->send_message2(m);
}


void MDSRank::send_message_mds(const ref_t<Message>& m, mds_rank_t mds)
{
  if (!mdsmap->is_up(mds)) {
    dout(10) << "send_message_mds mds." << mds << " not up, dropping " << *m << dendl;
    return;
  }

  // send mdsmap first?
  if (mds != whoami && peer_mdsmap_epoch[mds] < mdsmap->get_epoch()) {
    auto _m = make_message<MMDSMap>(monc->get_fsid(), *mdsmap);
    messenger->send_to_mds(_m.detach(), mdsmap->get_addrs(mds));
    peer_mdsmap_epoch[mds] = mdsmap->get_epoch();
  }

  // send message
  messenger->send_to_mds(ref_t<Message>(m).detach(), mdsmap->get_addrs(mds));
}

void MDSRank::forward_message_mds(const cref_t<MClientRequest>& m, mds_rank_t mds)
{
  ceph_assert(mds != whoami);

  /*
   * don't actually forward if non-idempotent!
   * client has to do it.  although the MDS will ignore duplicate requests,
   * the affected metadata may migrate, in which case the new authority
   * won't have the metareq_id in the completed request map.
   */
  // NEW: always make the client resend!
  bool client_must_resend = true;  //!creq->can_forward();

  // tell the client where it should go
  auto session = get_session(m);
  auto f = make_message<MClientRequestForward>(m->get_tid(), mds, m->get_num_fwd()+1, client_must_resend);
  send_message_client(f, session);
}

void MDSRank::send_message_client_counted(const ref_t<Message>& m, client_t client)
{
  Session *session = sessionmap.get_session(entity_name_t::CLIENT(client.v));
  if (session) {
    send_message_client_counted(m, session);
  } else {
    dout(10) << "send_message_client_counted no session for client." << client << " " << *m << dendl;
  }
}

void MDSRank::send_message_client_counted(const ref_t<Message>& m, const ConnectionRef& connection)
{
  // do not carry ref
  auto session = static_cast<Session *>(connection->get_priv().get());
  if (session) {
    send_message_client_counted(m, session);
  } else {
    dout(10) << "send_message_client_counted has no session for " << m->get_source_inst() << dendl;
    // another Connection took over the Session
  }
}

void MDSRank::send_message_client_counted(const ref_t<Message>& m, Session* session)
{
  version_t seq = session->inc_push_seq();
  dout(10) << "send_message_client_counted " << session->info.inst.name << " seq "
           << seq << " " << *m << dendl;
  if (session->get_connection()) {
    session->get_connection()->send_message2(m);
  } else {
    session->preopen_out_queue.push_back(m);
  }
}

void MDSRank::send_message_client(const ref_t<Message>& m, Session* session)
{
  dout(10) << "send_message_client " << session->info.inst << " " << *m << dendl;
  if (session->get_connection()) {
    session->get_connection()->send_message2(m);
  } else {
    session->preopen_out_queue.push_back(m);
  }
}

/**
 * This is used whenever a RADOS operation has been cancelled
 * or a RADOS client has been blacklisted, to cause the MDS and
 * any clients to wait for this OSD epoch before using any new caps.
 *
 * See doc/cephfs/eviction
 */
void MDSRank::set_osd_epoch_barrier(epoch_t e)
{
  dout(4) << __func__ << ": epoch=" << e << dendl;
  osd_epoch_barrier = e;
}

void MDSRank::retry_dispatch(const cref_t<Message> &m)
{
  inc_dispatch_depth();
  _dispatch(m, false);
  dec_dispatch_depth();
}

double MDSRank::get_dispatch_queue_max_age(utime_t now) const
{
  return messenger->get_dispatch_queue_max_age(now);
}

bool MDSRank::is_daemon_stopping() const
{
  return stopping;
}

void MDSRank::request_state(MDSMap::DaemonState s)
{
  dout(3) << "request_state " << ceph_mds_state_name(s) << dendl;
  beacon.set_want_state(*mdsmap, s);
  beacon.send();
}


class C_MDS_BootStart : public MDSInternalContext {
  MDSRank::BootStep nextstep;
public:
  C_MDS_BootStart(MDSRank *m, MDSRank::BootStep n)
    : MDSInternalContext(m), nextstep(n) {}
  void finish(int r) override {
    mds->boot_start(nextstep, r);
  }
};


void MDSRank::boot_start(BootStep step, int r)
{
  // Handle errors from previous step
  if (r < 0) {
    if (is_standby_replay() && (r == -EAGAIN)) {
      dout(0) << "boot_start encountered an error EAGAIN"
              << ", respawning since we fell behind journal" << dendl;
      respawn();
    } else if (r == -EINVAL || r == -ENOENT) {
      // Invalid or absent data, indicates damaged on-disk structures
      clog->error() << "Error loading MDS rank " << whoami << ": "
        << cpp_strerror(r);
      damaged();
      ceph_assert(r == 0);  // Unreachable, damaged() calls respawn()
    } else if (r == -EROFS) {
      dout(0) << "boot error forcing transition to read-only; MDS will try to continue" << dendl;
    } else {
      // Completely unexpected error, give up and die
      dout(0) << "boot_start encountered an error, failing" << dendl;
      suicide();
      return;
    }
  }

  ceph_assert(is_starting() || is_any_replay());

  switch(step) {
    case MDS_BOOT_INITIAL:
      {
        mdcache->init_layouts();

        MDSGatherBuilder gather(g_ceph_context,
            new C_MDS_BootStart(this, MDS_BOOT_OPEN_ROOT));
        dout(2) << "Booting: " << step << ": opening inotable" << dendl;
        inotable->set_rank(whoami);
        inotable->load(gather.new_sub());

        dout(2) << "Booting: " << step << ": opening sessionmap" << dendl;
        sessionmap.set_rank(whoami);
        sessionmap.load(gather.new_sub());

        dout(2) << "Booting: " << step << ": opening mds log" << dendl;
        mdlog->open(gather.new_sub());

        if (is_starting()) {
          dout(2) << "Booting: " << step << ": opening purge queue" << dendl;
          purge_queue.open(new C_IO_Wrapper(this, gather.new_sub()));
        } else if (!standby_replaying) {
          dout(2) << "Booting: " << step << ": opening purge queue (async)" << dendl;
          purge_queue.open(NULL);
          dout(2) << "Booting: " << step << ": loading open file table (async)" << dendl;
          mdcache->open_file_table.load(nullptr);
        }

        if (mdsmap->get_tableserver() == whoami) {
          dout(2) << "Booting: " << step << ": opening snap table" << dendl;
          snapserver->set_rank(whoami);
          snapserver->load(gather.new_sub());
        }

        gather.activate();
      }
      break;
    case MDS_BOOT_OPEN_ROOT:
      {
        dout(2) << "Booting: " << step << ": loading/discovering base inodes" << dendl;

        MDSGatherBuilder gather(g_ceph_context,
            new C_MDS_BootStart(this, MDS_BOOT_PREPARE_LOG));

        if (is_starting()) {
          // load mydir frag for the first log segment (creating subtree map)
          mdcache->open_mydir_frag(gather.new_sub());
        } else {
          mdcache->open_mydir_inode(gather.new_sub());
        }

        mdcache->create_global_snaprealm();

        if (whoami == mdsmap->get_root()) {  // load root inode off disk if we are auth
          mdcache->open_root_inode(gather.new_sub());
        } else if (is_any_replay()) {
          // replay.  make up fake root inode to start with
          mdcache->create_root_inode();
        }
        gather.activate();
      }
      break;
    case MDS_BOOT_PREPARE_LOG:
      if (is_any_replay()) {
        dout(2) << "Booting: " << step << ": replaying mds log" << dendl;
        MDSGatherBuilder gather(g_ceph_context,
            new C_MDS_BootStart(this, MDS_BOOT_REPLAY_DONE));

        if (!standby_replaying) {
          dout(2) << "Booting: " << step << ": waiting for purge queue recovered" << dendl;
          purge_queue.wait_for_recovery(new C_IO_Wrapper(this, gather.new_sub()));
        }

        mdlog->replay(gather.new_sub());
        gather.activate();
      } else {
        dout(2) << "Booting: " << step << ": positioning at end of old mds log" << dendl;
        mdlog->append();
        starting_done();
      }
      break;
    case MDS_BOOT_REPLAY_DONE:
      ceph_assert(is_any_replay());

      // Sessiontable and inotable should be in sync after replay, validate
      // that they are consistent.
      validate_sessions();

      replay_done();
      break;
  }
}

void MDSRank::validate_sessions()
{
  ceph_assert(ceph_mutex_is_locked_by_me(mds_lock));
  bool valid = true;

  // Identify any sessions which have state inconsistent with other,
  // after they have been loaded from rados during startup.
  // Mitigate bugs like: http://tracker.ceph.com/issues/16842
  for (const auto &i : sessionmap.get_sessions()) {
    Session *session = i.second;
    interval_set<inodeno_t> badones;
    if (inotable->intersects_free(session->info.prealloc_inos, &badones)) {
      clog->error() << "client " << *session
                    << "loaded with preallocated inodes that are inconsistent with inotable";
      valid = false;
    }
  }

  if (!valid) {
    damaged();
    ceph_assert(valid);
  }
}

void MDSRank::starting_done()
{
  dout(3) << "starting_done" << dendl;
  ceph_assert(is_starting());
  request_state(MDSMap::STATE_ACTIVE);

  mdlog->start_new_segment();

  // sync snaptable cache
  snapclient->sync(new C_MDSInternalNoop);
}


void MDSRank::calc_recovery_set()
{
  // initialize gather sets
  set<mds_rank_t> rs;
  mdsmap->get_recovery_mds_set(rs);
  rs.erase(whoami);
  mdcache->set_recovery_set(rs);

  dout(1) << " recovery set is " << rs << dendl;
}


void MDSRank::replay_start()
{
  dout(1) << "replay_start" << dendl;

  if (is_standby_replay())
    standby_replaying = true;

  calc_recovery_set();

  // Check if we need to wait for a newer OSD map before starting
  Context *fin = new C_IO_Wrapper(this, new C_MDS_BootStart(this, MDS_BOOT_INITIAL));
  bool const ready = objecter->wait_for_map(
      mdsmap->get_last_failure_osd_epoch(),
      fin);

  if (ready) {
    delete fin;
    boot_start();
  } else {
    dout(1) << " waiting for osdmap " << mdsmap->get_last_failure_osd_epoch()
            << " (which blacklists prior instance)" << dendl;
  }
}


class MDSRank::C_MDS_StandbyReplayRestartFinish : public MDSIOContext {
  uint64_t old_read_pos;
public:
  C_MDS_StandbyReplayRestartFinish(MDSRank *mds_, uint64_t old_read_pos_) :
    MDSIOContext(mds_), old_read_pos(old_read_pos_) {}
  void finish(int r) override {
    mds->_standby_replay_restart_finish(r, old_read_pos);
  }
  void print(ostream& out) const override {
    out << "standby_replay_restart";
  }
};

void MDSRank::_standby_replay_restart_finish(int r, uint64_t old_read_pos)
{
  if (old_read_pos < mdlog->get_journaler()->get_trimmed_pos()) {
    dout(0) << "standby MDS fell behind active MDS journal's expire_pos, restarting" << dendl;
    respawn(); /* we're too far back, and this is easier than
                  trying to reset everything in the cache, etc */
  } else {
    mdlog->standby_trim_segments();
    boot_start(MDS_BOOT_PREPARE_LOG, r);
  }
}

class MDSRank::C_MDS_StandbyReplayRestart : public MDSInternalContext {
public:
  explicit C_MDS_StandbyReplayRestart(MDSRank *m) : MDSInternalContext(m) {}
  void finish(int r) override {
    ceph_assert(!r);
    mds->standby_replay_restart();
  }
};

void MDSRank::standby_replay_restart()
{
  if (standby_replaying) {
    /* Go around for another pass of replaying in standby */
    dout(5) << "Restarting replay as standby-replay" << dendl;
    mdlog->get_journaler()->reread_head_and_probe(
      new C_MDS_StandbyReplayRestartFinish(
        this,
        mdlog->get_journaler()->get_read_pos()));
  } else {
    /* We are transitioning out of standby: wait for OSD map update
       before making final pass */
    dout(1) << "standby_replay_restart (final takeover pass)" << dendl;
    Context *fin = new C_IO_Wrapper(this, new C_MDS_StandbyReplayRestart(this));
    bool ready = objecter->wait_for_map(mdsmap->get_last_failure_osd_epoch(), fin);
    if (ready) {
      delete fin;
      mdlog->get_journaler()->reread_head_and_probe(
        new C_MDS_StandbyReplayRestartFinish(
          this,
          mdlog->get_journaler()->get_read_pos()));

      dout(1) << " opening purge_queue (async)" << dendl;
      purge_queue.open(NULL);
      dout(1) << " opening open_file_table (async)" << dendl;
      mdcache->open_file_table.load(nullptr);
    } else {
      dout(1) << " waiting for osdmap " << mdsmap->get_last_failure_osd_epoch()
              << " (which blacklists prior instance)" << dendl;
    }
  }
}

void MDSRank::replay_done()
{
  if (!standby_replaying) {
    dout(1) << "Finished replaying journal" << dendl;
  } else {
    dout(5) << "Finished replaying journal as standby-replay" << dendl;
  }

  if (is_standby_replay()) {
    // The replay was done in standby state, and we are still in that state
    ceph_assert(standby_replaying);
    dout(10) << "setting replay timer" << dendl;
    timer.add_event_after(g_conf()->mds_replay_interval,
                          new C_MDS_StandbyReplayRestart(this));
    return;
  } else if (standby_replaying) {
    // The replay was done in standby state, we have now _left_ that state
    dout(10) << " last replay pass was as a standby; making final pass" << dendl;
    standby_replaying = false;
    standby_replay_restart();
    return;
  } else {
    // Replay is complete, journal read should be up to date
    ceph_assert(mdlog->get_journaler()->get_read_pos() == mdlog->get_journaler()->get_write_pos());
    ceph_assert(!is_standby_replay());

    // Reformat and come back here
    if (mdlog->get_journaler()->get_stream_format() < g_conf()->mds_journal_format) {
        dout(4) << "reformatting journal on standby-replay->replay transition" << dendl;
        mdlog->reopen(new C_MDS_BootStart(this, MDS_BOOT_REPLAY_DONE));
        return;
    }
  }

  dout(1) << "making mds journal writeable" << dendl;
  mdlog->get_journaler()->set_writeable();
  mdlog->get_journaler()->trim_tail();

  if (mdsmap->get_tableserver() == whoami &&
      snapserver->upgrade_format()) {
    dout(1) << "upgrading snaptable format" << dendl;
    snapserver->save(new C_MDSInternalNoop);
  }

  if (g_conf()->mds_wipe_sessions) {
    dout(1) << "wiping out client sessions" << dendl;
    sessionmap.wipe();
    sessionmap.save(new C_MDSInternalNoop);
  }
  if (g_conf()->mds_wipe_ino_prealloc) {
    dout(1) << "wiping out ino prealloc from sessions" << dendl;
    sessionmap.wipe_ino_prealloc();
    sessionmap.save(new C_MDSInternalNoop);
  }
  if (g_conf()->mds_skip_ino) {
    inodeno_t i = g_conf()->mds_skip_ino;
    dout(1) << "skipping " << i << " inodes" << dendl;
    inotable->skip_inos(i);
    inotable->save(new C_MDSInternalNoop);
  }

  if (mdsmap->get_num_in_mds() == 1 &&
      mdsmap->get_num_failed_mds() == 0) { // just me!
    dout(2) << "i am alone, moving to state reconnect" << dendl;
    request_state(MDSMap::STATE_RECONNECT);
    // sync snaptable cache
    snapclient->sync(new C_MDSInternalNoop);
  } else {
    dout(2) << "i am not alone, moving to state resolve" << dendl;
    request_state(MDSMap::STATE_RESOLVE);
  }
}

void MDSRank::reopen_log()
{
  dout(1) << "reopen_log" << dendl;
  mdcache->rollback_uncommitted_fragments();
}

void MDSRank::resolve_start()
{
  dout(1) << "resolve_start" << dendl;

  reopen_log();

  mdcache->resolve_start(new C_MDS_VoidFn(this, &MDSRank::resolve_done));
  finish_contexts(g_ceph_context, waiting_for_resolve);
}

void MDSRank::resolve_done()
{
  dout(1) << "resolve_done" << dendl;
  request_state(MDSMap::STATE_RECONNECT);
  // sync snaptable cache
  snapclient->sync(new C_MDSInternalNoop);
}

void MDSRank::reconnect_start()
{
  dout(1) << "reconnect_start" << dendl;

  if (last_state == MDSMap::STATE_REPLAY) {
    reopen_log();
  }

  // Drop any blacklisted clients from the SessionMap before going
  // into reconnect, so that we don't wait for them.
  objecter->enable_blacklist_events();
  std::set<entity_addr_t> blacklist;
  epoch_t epoch = 0;
  objecter->with_osdmap([&blacklist, &epoch](const OSDMap& o) {
      o.get_blacklist(&blacklist);
      epoch = o.get_epoch();
  });
  auto killed = server->apply_blacklist(blacklist);
  dout(4) << "reconnect_start: killed " << killed << " blacklisted sessions ("
          << blacklist.size() << " blacklist entries, "
          << sessionmap.get_sessions().size() << ")" << dendl;
  if (killed) {
    set_osd_epoch_barrier(epoch);
  }

  server->reconnect_clients(new C_MDS_VoidFn(this, &MDSRank::reconnect_done));
  finish_contexts(g_ceph_context, waiting_for_reconnect);
}
void MDSRank::reconnect_done()
{
  dout(1) << "reconnect_done" << dendl;
  request_state(MDSMap::STATE_REJOIN);    // move to rejoin state
}

void MDSRank::rejoin_joint_start()
{
  dout(1) << "rejoin_joint_start" << dendl;
  mdcache->rejoin_send_rejoins();
}
void MDSRank::rejoin_start()
{
  dout(1) << "rejoin_start" << dendl;
  mdcache->rejoin_start(new C_MDS_VoidFn(this, &MDSRank::rejoin_done));
  finish_contexts(g_ceph_context, waiting_for_rejoin);
}
void MDSRank::rejoin_done()
{
  dout(1) << "rejoin_done" << dendl;
  mdcache->show_subtrees();
  mdcache->show_cache();

  if (mdcache->is_any_uncommitted_fragment()) {
    dout(1) << " waiting for uncommitted fragments" << dendl;
    MDSGatherBuilder gather(g_ceph_context, new C_MDS_VoidFn(this, &MDSRank::rejoin_done));
    mdcache->wait_for_uncommitted_fragments(gather.get());
    gather.activate();
    return;
  }

  // funny case: is our cache empty?  no subtrees?
  if (!mdcache->is_subtrees()) {
    if (whoami == 0) {
      // The root should always have a subtree!
      clog->error() << "No subtrees found for root MDS rank!";
      damaged();
      ceph_assert(mdcache->is_subtrees());
    } else {
      dout(1) << " empty cache, no subtrees, leaving cluster" << dendl;
      request_state(MDSMap::STATE_STOPPED);
    }
    return;
  }

  if (replay_queue.empty() && !server->get_num_pending_reclaim()) {
    request_state(MDSMap::STATE_ACTIVE);
  } else {
    replaying_requests_done = replay_queue.empty();
    request_state(MDSMap::STATE_CLIENTREPLAY);
  }
}

void MDSRank::clientreplay_start()
{
  dout(1) << "clientreplay_start" << dendl;
  finish_contexts(g_ceph_context, waiting_for_replay);  // kick waiters
  mdcache->start_files_to_recover();
  queue_one_replay();
}

bool MDSRank::queue_one_replay()
{
  if (!replay_queue.empty()) {
    queue_waiter(replay_queue.front());
    replay_queue.pop_front();
    return true;
  }
  if (!replaying_requests_done) {
    replaying_requests_done = true;
    mdlog->flush();
  }
  maybe_clientreplay_done();
  return false;
}

void MDSRank::maybe_clientreplay_done()
{
  if (is_clientreplay() && get_want_state() == MDSMap::STATE_CLIENTREPLAY) {

    // don't go to active if there are session waiting for being reclaimed
    if (replaying_requests_done && !server->get_num_pending_reclaim()) {
      mdlog->wait_for_safe(new C_MDS_VoidFn(this, &MDSRank::clientreplay_done));
      return;
    }

    dout(1) << " still have " << replay_queue.size() + (int)!replaying_requests_done
            << " requests need to be replayed, " << server->get_num_pending_reclaim()
            << " sessions need to be reclaimed" << dendl;
  }
}

void MDSRank::clientreplay_done()
{
  dout(1) << "clientreplay_done" << dendl;
  request_state(MDSMap::STATE_ACTIVE);
}

void MDSRank::active_start()
{
  dout(1) << "active_start" << dendl;

  if (last_state == MDSMap::STATE_CREATING ||
      last_state == MDSMap::STATE_STARTING) {
    mdcache->open_root();
  }

  mdcache->clean_open_file_lists();
  mdcache->export_remaining_imported_caps();
  finish_contexts(g_ceph_context, waiting_for_replay);  // kick waiters
  mdcache->start_files_to_recover();

  mdcache->reissue_all_caps();

  finish_contexts(g_ceph_context, waiting_for_active);  // kick waiters
}

void MDSRank::recovery_done(int oldstate)
{
  dout(1) << "recovery_done -- successful recovery!" << dendl;
  ceph_assert(is_clientreplay() || is_active());

  if (oldstate == MDSMap::STATE_CREATING)
    return;

  mdcache->start_recovered_truncates();
  mdcache->start_purge_inodes();
  mdcache->do_file_recover();

  // tell connected clients
  //bcast_mds_map();     // not anymore, they get this from the monitor

  mdcache->populate_mydir();
}

void MDSRank::creating_done()
{
  dout(1)<< "creating_done" << dendl;
  request_state(MDSMap::STATE_ACTIVE);
  // sync snaptable cache
  snapclient->sync(new C_MDSInternalNoop);
}

void MDSRank::boot_create()
{
  dout(3) << "boot_create" << dendl;

  MDSGatherBuilder fin(g_ceph_context, new C_MDS_VoidFn(this, &MDSRank::creating_done));

  mdcache->init_layouts();

  inotable->set_rank(whoami);
  sessionmap.set_rank(whoami);

  // start with a fresh journal
  dout(10) << "boot_create creating fresh journal" << dendl;
  mdlog->create(fin.new_sub());

  // open new journal segment, but do not journal subtree map (yet)
  mdlog->prepare_new_segment();

  if (whoami == mdsmap->get_root()) {
    dout(3) << "boot_create creating fresh hierarchy" << dendl;
    mdcache->create_empty_hierarchy(fin.get());
  }

  dout(3) << "boot_create creating mydir hierarchy" << dendl;
  mdcache->create_mydir_hierarchy(fin.get());

  dout(3) << "boot_create creating global snaprealm" << dendl;
  mdcache->create_global_snaprealm();

  // fixme: fake out inotable (reset, pretend loaded)
  dout(10) << "boot_create creating fresh inotable table" << dendl;
  inotable->reset();
  inotable->save(fin.new_sub());

  // write empty sessionmap
  sessionmap.save(fin.new_sub());

  // Create empty purge queue
  purge_queue.create(new C_IO_Wrapper(this, fin.new_sub()));

  // initialize tables
  if (mdsmap->get_tableserver() == whoami) {
    dout(10) << "boot_create creating fresh snaptable" << dendl;
    snapserver->set_rank(whoami);
    snapserver->reset();
    snapserver->save(fin.new_sub());
  }

  ceph_assert(g_conf()->mds_kill_create_at != 1);

  // ok now journal it
  mdlog->journal_segment_subtree_map(fin.new_sub());
  mdlog->flush();

  // Usually we do this during reconnect, but creation skips that.
  objecter->enable_blacklist_events();

  fin.activate();
}

void MDSRank::stopping_start()
{
  dout(2) << "Stopping..." << dendl;

  if (mdsmap->get_num_in_mds() == 1 && !sessionmap.empty()) {
    std::vector<Session*> victims;
    const auto& sessions = sessionmap.get_sessions();
    for (const auto& p : sessions)  {
      if (!p.first.is_client()) {
        continue;
      }

      Session *s = p.second;
      victims.push_back(s);
    }

    dout(20) << __func__ << " matched " << victims.size() << " sessions" << dendl;
    ceph_assert(!victims.empty());

    C_GatherBuilder gather(g_ceph_context, new C_MDSInternalNoop);
    for (const auto &s : victims) {
      std::stringstream ss;
      evict_client(s->get_client().v, false,
                   g_conf()->mds_session_blacklist_on_evict, ss, gather.new_sub());
    }
    gather.activate();
  }

  mdcache->shutdown_start();
}

void MDSRank::stopping_done()
{
  dout(2) << "Finished stopping..." << dendl;

  // tell monitor we shut down cleanly.
  request_state(MDSMap::STATE_STOPPED);
}

void MDSRankDispatcher::handle_mds_map(
    const cref_t<MMDSMap> &m,
    const MDSMap &oldmap)
{
  // I am only to be passed MDSMaps in which I hold a rank
  ceph_assert(whoami != MDS_RANK_NONE);

  MDSMap::DaemonState oldstate = state;
  mds_gid_t mds_gid = mds_gid_t(monc->get_global_id());
  state = mdsmap->get_state_gid(mds_gid);
  if (state != oldstate) {
    last_state = oldstate;
    incarnation = mdsmap->get_inc_gid(mds_gid);
  }

  version_t epoch = m->get_epoch();

  // note source's map version
  if (m->get_source().is_mds() &&
      peer_mdsmap_epoch[mds_rank_t(m->get_source().num())] < epoch) {
    dout(15) << " peer " << m->get_source()
             << " has mdsmap epoch >= " << epoch
             << dendl;
    peer_mdsmap_epoch[mds_rank_t(m->get_source().num())] = epoch;
  }

  // Validate state transitions while I hold a rank
  if (!MDSMap::state_transition_valid(oldstate, state)) {
    derr << "Invalid state transition " << ceph_mds_state_name(oldstate)
      << "->" << ceph_mds_state_name(state) << dendl;
    respawn();
  }

  if (oldstate != state) {
    // update messenger.
    if (state == MDSMap::STATE_STANDBY_REPLAY) {
      dout(1) << "handle_mds_map i am now mds." << mds_gid << "." << incarnation
              << " replaying mds." << whoami << "." << incarnation << dendl;
      messenger->set_myname(entity_name_t::MDS(mds_gid));
    } else {
      dout(1) << "handle_mds_map i am now mds." << whoami << "." << incarnation << dendl;
      messenger->set_myname(entity_name_t::MDS(whoami));
    }
  }

  // tell objecter my incarnation
  if (objecter->get_client_incarnation() != incarnation)
    objecter->set_client_incarnation(incarnation);

  if (mdsmap->get_min_compat_client() < ceph_release_t::max &&
      oldmap.get_min_compat_client() != mdsmap->get_min_compat_client())
    server->update_required_client_features();

  // for debug
  if (g_conf()->mds_dump_cache_on_map)
    mdcache->dump_cache();

  cluster_degraded = mdsmap->is_degraded();

  // mdsmap and oldmap can be discontinuous. failover might happen in the missing mdsmap.
  // the 'restart' set tracks ranks that have restarted since the old mdsmap
  set<mds_rank_t> restart;
  // replaying mds does not communicate with other ranks
  if (state >= MDSMap::STATE_RESOLVE) {
    // did someone fail?
    //   new down?
    set<mds_rank_t> olddown, down;
    oldmap.get_down_mds_set(&olddown);
    mdsmap->get_down_mds_set(&down);
    for (const auto& r : down) {
      if (oldmap.have_inst(r) && olddown.count(r) == 0) {
        messenger->mark_down_addrs(oldmap.get_addrs(r));
        handle_mds_failure(r);
      }
    }

    // did someone fail?
    //   did their addr/inst change?
    set<mds_rank_t> up;
    mdsmap->get_up_mds_set(up);
    for (const auto& r : up) {
      auto& info = mdsmap->get_info(r);
      if (oldmap.have_inst(r)) {
        auto& oldinfo = oldmap.get_info(r);
        if (info.inc != oldinfo.inc) {
          messenger->mark_down_addrs(oldinfo.get_addrs());
          if (info.state == MDSMap::STATE_REPLAY ||
              info.state == MDSMap::STATE_RESOLVE) {
            restart.insert(r);
            handle_mds_failure(r);
          } else {
            ceph_assert(info.state == MDSMap::STATE_STARTING ||
                   info.state == MDSMap::STATE_ACTIVE);
            // -> stopped (missing) -> starting -> active
            restart.insert(r);
            mdcache->migrator->handle_mds_failure_or_stop(r);
            if (mdsmap->get_tableserver() == whoami)
              snapserver->handle_mds_failure_or_stop(r);
          }
        }
      } else {
        if (info.state == MDSMap::STATE_REPLAY ||
            info.state == MDSMap::STATE_RESOLVE) {
          // -> starting/creating (missing) -> active (missing) -> replay -> resolve
          restart.insert(r);
          handle_mds_failure(r);
        } else {
          ceph_assert(info.state == MDSMap::STATE_CREATING ||
                 info.state == MDSMap::STATE_STARTING ||
                 info.state == MDSMap::STATE_ACTIVE);
        }
      }
    }
  }

  // did it change?
  if (oldstate != state) {
    dout(1) << "handle_mds_map state change "
            << ceph_mds_state_name(oldstate) << " --> "
            << ceph_mds_state_name(state) << dendl;
    beacon.set_want_state(*mdsmap, state);

    if (oldstate == MDSMap::STATE_STANDBY_REPLAY) {
        dout(10) << "Monitor activated us! Deactivating replay loop" << dendl;
        assert (state == MDSMap::STATE_REPLAY);
    } else {
      // did i just recover?
      if ((is_active() || is_clientreplay()) &&
          (oldstate == MDSMap::STATE_CREATING ||
           oldstate == MDSMap::STATE_REJOIN ||
           oldstate == MDSMap::STATE_RECONNECT))
        recovery_done(oldstate);

      if (is_active()) {
        active_start();
      } else if (is_any_replay()) {
        replay_start();
      } else if (is_resolve()) {
        resolve_start();
      } else if (is_reconnect()) {
        reconnect_start();
      } else if (is_rejoin()) {
        rejoin_start();
      } else if (is_clientreplay()) {
        clientreplay_start();
      } else if (is_creating()) {
        boot_create();
      } else if (is_starting()) {
        boot_start();
      } else if (is_stopping()) {
        ceph_assert(oldstate == MDSMap::STATE_ACTIVE);
        stopping_start();
      }
    }
  }

  // RESOLVE
  // is someone else newly resolving?
  if (state >= MDSMap::STATE_RESOLVE) {
    // recover snaptable
    if (mdsmap->get_tableserver() == whoami) {
      if (oldstate < MDSMap::STATE_RESOLVE) {
        set<mds_rank_t> s;
        mdsmap->get_mds_set_lower_bound(s, MDSMap::STATE_RESOLVE);
        snapserver->finish_recovery(s);
      } else {
        set<mds_rank_t> old_set, new_set;
        oldmap.get_mds_set_lower_bound(old_set, MDSMap::STATE_RESOLVE);
        mdsmap->get_mds_set_lower_bound(new_set, MDSMap::STATE_RESOLVE);
        for (const auto& r : new_set) {
          if (r == whoami)
            continue; // not me
          if (!old_set.count(r) || restart.count(r)) {  // newly so?
            snapserver->handle_mds_recovery(r);
          }
        }
      }
    }

    if ((!oldmap.is_resolving() || !restart.empty()) && mdsmap->is_resolving()) {
      set<mds_rank_t> resolve;
      mdsmap->get_mds_set(resolve, MDSMap::STATE_RESOLVE);
      dout(10) << " resolve set is " << resolve << dendl;
      calc_recovery_set();
      mdcache->send_resolves();
    }
  }

  // REJOIN
  // is everybody finally rejoining?
  if (state >= MDSMap::STATE_REJOIN) {
    // did we start?
    if (!oldmap.is_rejoining() && mdsmap->is_rejoining())
      rejoin_joint_start();

    // did we finish?
    if (g_conf()->mds_dump_cache_after_rejoin &&
        oldmap.is_rejoining() && !mdsmap->is_rejoining())
      mdcache->dump_cache();      // for DEBUG only

    if (oldstate >= MDSMap::STATE_REJOIN ||
        oldstate == MDSMap::STATE_STARTING) {
      // ACTIVE|CLIENTREPLAY|REJOIN => we can discover from them.
      set<mds_rank_t> olddis, dis;
      oldmap.get_mds_set_lower_bound(olddis, MDSMap::STATE_REJOIN);
      mdsmap->get_mds_set_lower_bound(dis, MDSMap::STATE_REJOIN);
      for (const auto& r : dis) {
        if (r == whoami)
          continue; // not me
        if (!olddis.count(r) || restart.count(r)) {  // newly so?
          mdcache->kick_discovers(r);
          mdcache->kick_open_ino_peers(r);
        }
      }
    }
  }

  if (oldmap.is_degraded() && !cluster_degraded && state >= MDSMap::STATE_ACTIVE) {
    dout(1) << "cluster recovered." << dendl;
    auto it = waiting_for_active_peer.find(MDS_RANK_NONE);
    if (it != waiting_for_active_peer.end()) {
      queue_waiters(it->second);
      waiting_for_active_peer.erase(it);
    }
  }

  // did someone go active?
  if (state >= MDSMap::STATE_CLIENTREPLAY &&
      oldstate >= MDSMap::STATE_CLIENTREPLAY) {
    set<mds_rank_t> oldactive, active;
    oldmap.get_mds_set_lower_bound(oldactive, MDSMap::STATE_CLIENTREPLAY);
    mdsmap->get_mds_set_lower_bound(active, MDSMap::STATE_CLIENTREPLAY);
    for (const auto& r : active) {
      if (r == whoami)
        continue; // not me
      if (!oldactive.count(r) || restart.count(r))  // newly so?
        handle_mds_recovery(r);
    }
  }

  if (is_clientreplay() || is_active() || is_stopping()) {
    // did anyone stop?
    set<mds_rank_t> oldstopped, stopped;
    oldmap.get_stopped_mds_set(oldstopped);
    mdsmap->get_stopped_mds_set(stopped);
    for (const auto& r : stopped)
      if (oldstopped.count(r) == 0) {     // newly so?
        mdcache->migrator->handle_mds_failure_or_stop(r);
        if (mdsmap->get_tableserver() == whoami)
          snapserver->handle_mds_failure_or_stop(r);
      }
  }

  {
    map<epoch_t,MDSContext::vec >::iterator p = waiting_for_mdsmap.begin();
    while (p != waiting_for_mdsmap.end() && p->first <= mdsmap->get_epoch()) {
      MDSContext::vec ls;
      ls.swap(p->second);
      waiting_for_mdsmap.erase(p++);
      queue_waiters(ls);
    }
  }

  if (is_active()) {
    // Before going active, set OSD epoch barrier to latest (so that
    // we don't risk handing out caps to clients with old OSD maps that
    // might not include barriers from the previous incarnation of this MDS)
    set_osd_epoch_barrier(objecter->with_osdmap(
                            std::mem_fn(&OSDMap::get_epoch)));

    /* Now check if we should hint to the OSD that a read may follow */
    if (mdsmap->has_standby_replay(whoami))
      mdlog->set_write_iohint(0);
    else
      mdlog->set_write_iohint(CEPH_OSD_OP_FLAG_FADVISE_DONTNEED);
  }

  if (oldmap.get_max_mds() != mdsmap->get_max_mds()) {
    purge_queue.update_op_limit(*mdsmap);
  }

  if (mdsmap->get_inline_data_enabled() && !oldmap.get_inline_data_enabled())
    dout(0) << "WARNING: inline_data support has been deprecated and will be removed in a future release" << dendl;

  if (scrubstack->is_scrubbing()) {
    if (mdsmap->get_max_mds() > 1) {
      auto c = new C_MDSInternalNoop;
      scrubstack->scrub_abort(c);
    }
  }
  mdcache->handle_mdsmap(*mdsmap, oldmap);
}

void MDSRank::handle_mds_recovery(mds_rank_t who)
{
  dout(5) << "handle_mds_recovery mds." << who << dendl;

  mdcache->handle_mds_recovery(who);

  queue_waiters(waiting_for_active_peer[who]);
  waiting_for_active_peer.erase(who);
}

void MDSRank::handle_mds_failure(mds_rank_t who)
{
  if (who == whoami) {
    dout(5) << "handle_mds_failure for myself; not doing anything" << dendl;
    return;
  }
  dout(5) << "handle_mds_failure mds." << who << dendl;

  mdcache->handle_mds_failure(who);

  if (mdsmap->get_tableserver() == whoami)
    snapserver->handle_mds_failure_or_stop(who);

  snapclient->handle_mds_failure(who);
}

void MDSRankDispatcher::handle_asok_command(
  std::string_view command,
  const cmdmap_t& cmdmap,
  Formatter *f,
  const bufferlist &inbl,
  std::function<void(int,const std::string&,bufferlist&)> on_finish)
{
  int r = 0;
  stringstream ss;
  bufferlist outbl;
  if (command == "dump_ops_in_flight" ||
      command == "ops") {
    if (!op_tracker.dump_ops_in_flight(f)) {
      ss << "op_tracker disabled; set mds_enable_op_tracker=true to enable";
    }
  } else if (command == "dump_blocked_ops") {
    if (!op_tracker.dump_ops_in_flight(f, true)) {
      ss << "op_tracker disabled; set mds_enable_op_tracker=true to enable";
    }
  } else if (command == "dump_historic_ops") {
    if (!op_tracker.dump_historic_ops(f)) {
      ss << "op_tracker disabled; set mds_enable_op_tracker=true to enable";
    }
  } else if (command == "dump_historic_ops_by_duration") {
    if (!op_tracker.dump_historic_ops(f, true)) {
      ss << "op_tracker disabled; set mds_enable_op_tracker=true to enable";
    }
  } else if (command == "osdmap barrier") {
    int64_t target_epoch = 0;
    bool got_val = cmd_getval(cmdmap, "target_epoch", target_epoch);

    if (!got_val) {
      ss << "no target epoch given";
      r = -EINVAL;
      goto out;
    }
    {
      std::lock_guard l(mds_lock);
      set_osd_epoch_barrier(target_epoch);
    }
    C_SaferCond cond;
    bool already_got = objecter->wait_for_map(target_epoch, &cond);
    if (!already_got) {
      dout(4) << __func__ << ": waiting for OSD epoch " << target_epoch << dendl;
      cond.wait();
    }
  } else if (command == "session ls" ||
             command == "client ls") {
    std::lock_guard l(mds_lock);
    std::vector<std::string> filter_args;
    cmd_getval(cmdmap, "filters", filter_args);
    SessionFilter filter;
    r = filter.parse(filter_args, &ss);
    if (r != 0) {
      goto out;
    }
    dump_sessions(filter, f);
  } else if (command == "session evict" ||
             command == "client evict") {
    std::lock_guard l(mds_lock);
    std::vector<std::string> filter_args;
    cmd_getval(cmdmap, "filters", filter_args);

    SessionFilter filter;
    r = filter.parse(filter_args, &ss);
    if (r != 0) {
      r = -EINVAL;
      goto out;
    }
    evict_clients(filter, on_finish);
    return;
  } else if (command == "session kill") {
    std::string client_id;
    if (!cmd_getval(cmdmap, "client_id", client_id)) {
      ss << "Invalid client_id specified";
      r = -ENOENT;
      goto out;
    }
    std::lock_guard l(mds_lock);
    bool evicted = evict_client(strtol(client_id.c_str(), 0, 10), true,
        g_conf()->mds_session_blacklist_on_evict, ss);
    if (!evicted) {
      dout(15) << ss.str() << dendl;
      r = -ENOENT;
    }
  } else if (command == "session config" ||
             command == "client config") {
    int64_t client_id;
    std::string option;
    std::string value;

    cmd_getval(cmdmap, "client_id", client_id);
    cmd_getval(cmdmap, "option", option);
    bool got_value = cmd_getval(cmdmap, "value", value);

    std::lock_guard l(mds_lock);
    r = config_client(client_id, !got_value, option, value, ss);
  } else if (command == "scrub start" ||
             command == "scrub_start") {
    string path;
    string tag;
    vector<string> scrubop_vec;
    cmd_getval(cmdmap, "scrubops", scrubop_vec);
    cmd_getval(cmdmap, "path", path);
    cmd_getval(cmdmap, "tag", tag);

    /* Multiple MDS scrub is not currently supported. See also: https://tracker.ceph.com/issues/12274 */
    if (mdsmap->get_max_mds() > 1) {
      ss << "Scrub is not currently supported for multiple active MDS. Please reduce max_mds to 1 and then scrub.";
      r = -EINVAL;
      goto out;
    }

    finisher->queue(
      new LambdaContext(
        [this, on_finish, f, path, tag, scrubop_vec](int r) {
          command_scrub_start(
            f, path, tag, scrubop_vec,
            new LambdaContext(
              [on_finish](int r) {
                bufferlist outbl;
                on_finish(r, {}, outbl);
              }));
        }));
    return;
  } else if (command == "scrub abort") {
    finisher->queue(
      new LambdaContext(
        [this, on_finish, f](int r) {
          command_scrub_abort(
            f,
            new LambdaContext(
              [on_finish, f](int r) {
                bufferlist outbl;
                f->open_object_section("result");
                f->dump_int("return_code", r);
                f->close_section();
                on_finish(r, {}, outbl);
              }));
        }));
    return;
  } else if (command == "scrub pause") {
    finisher->queue(
      new LambdaContext(
        [this, on_finish, f](int r) {
          command_scrub_pause(
            f,
            new LambdaContext(
              [on_finish, f](int r) {
                bufferlist outbl;
                f->open_object_section("result");
                f->dump_int("return_code", r);
                f->close_section();
                on_finish(r, {}, outbl);
              }));
        }));
    return;
  } else if (command == "scrub resume") {
    command_scrub_resume(f);
  } else if (command == "scrub status") {
    command_scrub_status(f);
  } else if (command == "tag path") {
    string path;
    cmd_getval(cmdmap, "path", path);
    string tag;
    cmd_getval(cmdmap, "tag", tag);
    command_tag_path(f, path, tag);
  } else if (command == "flush_path") {
    string path;
    cmd_getval(cmdmap, "path", path);
    command_flush_path(f, path);
  } else if (command == "flush journal") {
    command_flush_journal(f);
  } else if (command == "get subtrees") {
    command_get_subtrees(f);
  } else if (command == "export dir") {
    string path;
    if(!cmd_getval(cmdmap, "path", path)) {
      ss << "malformed path";
      r = -EINVAL;
      goto out;
    }
    int64_t rank;
    if(!cmd_getval(cmdmap, "rank", rank)) {
      ss << "malformed rank";
      r = -EINVAL;
      goto out;
    }
    command_export_dir(f, path, (mds_rank_t)rank);
  } else if (command == "dump cache") {
    std::lock_guard l(mds_lock);
    string path;
    if (!cmd_getval(cmdmap, "path", path)) {
      r = mdcache->dump_cache(f);
    } else {
      r = mdcache->dump_cache(path);
    }
  } else if (command == "cache drop") {
    int64_t timeout = 0;
    cmd_getval(cmdmap, "timeout", timeout);
    finisher->queue(
      new LambdaContext(
        [this, on_finish, f, timeout](int r) {
          command_cache_drop(
            timeout, f,
            new LambdaContext(
              [on_finish](int r) {
                bufferlist outbl;
                on_finish(r, {}, outbl);
              }));
        }));
    return;
  } else if (command == "cache status") {
    std::lock_guard l(mds_lock);
    mdcache->cache_status(f);
  } else if (command == "dump tree") {
    command_dump_tree(cmdmap, ss, f);
  } else if (command == "dump loads") {
    std::lock_guard l(mds_lock);
    r = balancer->dump_loads(f);
  } else if (command == "dump snaps") {
    std::lock_guard l(mds_lock);
    string server;
    cmd_getval(cmdmap, "server", server);
    if (server == "--server") {
      if (mdsmap->get_tableserver() == whoami) {
        snapserver->dump(f);
      } else {
        r = -EXDEV;
        ss << "Not snapserver";
      }
    } else {
      r = snapclient->dump_cache(f);
    }
  } else if (command == "force_readonly") {
    std::lock_guard l(mds_lock);
    mdcache->force_readonly();
  } else if (command == "dirfrag split") {
    command_dirfrag_split(cmdmap, ss);
  } else if (command == "dirfrag merge") {
    command_dirfrag_merge(cmdmap, ss);
  } else if (command == "dirfrag ls") {
    command_dirfrag_ls(cmdmap, ss, f);
  } else if (command == "openfiles ls") {
    command_openfiles_ls(f);
  } else if (command == "dump inode") {
    command_dump_inode(f, cmdmap, ss);
  } else if (command == "damage ls") {
    std::lock_guard l(mds_lock);
    damage_table.dump(f);
  } else if (command == "damage rm") {
    std::lock_guard l(mds_lock);
    damage_entry_id_t id = 0;
    if (!cmd_getval(cmdmap, "damage_id", (int64_t&)id)) {
      r = -EINVAL;
      goto out;
    }
    damage_table.erase(id);
  } else {
    r = -ENOSYS;
  }
out:
  on_finish(r, ss.str(), outbl);
}

/**
 * This function drops the mds_lock, so don't do anything with
 * MDSRank after calling it (we could have gone into shutdown): just
 * send your result back to the calling client and finish.
 */
void MDSRankDispatcher::evict_clients(
  const SessionFilter &filter,
  std::function<void(int,const std::string&,bufferlist&)> on_finish)
{
  bufferlist outbl;
  if (is_any_replay()) {
    on_finish(-EAGAIN, "MDS is replaying log", outbl);
    return;
  }

  std::vector<Session*> victims;
  const auto& sessions = sessionmap.get_sessions();
  for (const auto& p : sessions)  {
    if (!p.first.is_client()) {
      continue;
    }

    Session *s = p.second;

    if (filter.match(*s, std::bind(&Server::waiting_for_reconnect, server,
                                   std::placeholders::_1))) {
      victims.push_back(s);
    }
  }

  dout(20) << __func__ << " matched " << victims.size() << " sessions" << dendl;

  if (victims.empty()) {
    on_finish(0, {}, outbl);
    return;
  }

  C_GatherBuilder gather(g_ceph_context,
                         new LambdaContext([on_finish](int r) {
                                             bufferlist bl;
                                             on_finish(r, {}, bl);
                                           }));
  for (const auto s : victims) {
    std::stringstream ss;
    evict_client(s->get_client().v, false,
                 g_conf()->mds_session_blacklist_on_evict, ss, gather.new_sub());
  }
  gather.activate();
}

void MDSRankDispatcher::dump_sessions(const SessionFilter &filter, Formatter *f) const
{
  // Dump sessions, decorated with recovery/replay status
  f->open_array_section("sessions");
  for (auto& [name, s] : sessionmap.get_sessions()) {
    if (!name.is_client()) {
      continue;
    }

    if (!filter.match(*s, std::bind(&Server::waiting_for_reconnect, server, std::placeholders::_1))) {
      continue;
    }

    f->dump_object("session", *s);
  }
  f->close_section(); // sessions
}

void MDSRank::command_scrub_start(Formatter *f,
                                  std::string_view path, std::string_view tag,
                                  const vector<string>& scrubop_vec, Context *on_finish)
{
  bool force = false;
  bool recursive = false;
  bool repair = false;
  for (auto &op : scrubop_vec) {
    if (op == "force")
      force = true;
    else if (op == "recursive")
      recursive = true;
    else if (op == "repair")
      repair = true;
  }

  std::lock_guard l(mds_lock);
  mdcache->enqueue_scrub(path, tag, force, recursive, repair, f, on_finish);
  // scrub_dentry() finishers will dump the data for us; we're done!
}

void MDSRank::command_tag_path(Formatter *f,
    std::string_view path, std::string_view tag)
{
  C_SaferCond scond;
  {
    std::lock_guard l(mds_lock);
    mdcache->enqueue_scrub(path, tag, true, true, false, f, &scond);
  }
  scond.wait();
}

void MDSRank::command_scrub_abort(Formatter *f, Context *on_finish) {
  std::lock_guard l(mds_lock);
  scrubstack->scrub_abort(on_finish);
}

void MDSRank::command_scrub_pause(Formatter *f, Context *on_finish) {
  std::lock_guard l(mds_lock);
  scrubstack->scrub_pause(on_finish);
}

void MDSRank::command_scrub_resume(Formatter *f) {
  std::lock_guard l(mds_lock);
  int r = scrubstack->scrub_resume();

  f->open_object_section("result");
  f->dump_int("return_code", r);
  f->close_section();
}

void MDSRank::command_scrub_status(Formatter *f) {
  std::lock_guard l(mds_lock);
  scrubstack->scrub_status(f);
}

void MDSRank::command_flush_path(Formatter *f, std::string_view path)
{
  C_SaferCond scond;
  {
    std::lock_guard l(mds_lock);
    mdcache->flush_dentry(path, &scond);
  }
  int r = scond.wait();
  f->open_object_section("results");
  f->dump_int("return_code", r);
  f->close_section(); // results
}

// synchronous wrapper around "journal flush" asynchronous context
// execution.
void MDSRank::command_flush_journal(Formatter *f) {
  ceph_assert(f != NULL);

  C_SaferCond cond;
  std::stringstream ss;
  {
    std::lock_guard locker(mds_lock);
    C_Flush_Journal *flush_journal = new C_Flush_Journal(mdcache, mdlog, this, &ss, &cond);
    flush_journal->send();
  }
  int r = cond.wait();

  f->open_object_section("result");
  f->dump_string("message", ss.str());
  f->dump_int("return_code", r);
  f->close_section();
}

void MDSRank::command_get_subtrees(Formatter *f)
{
  ceph_assert(f != NULL);
  std::lock_guard l(mds_lock);

  std::vector<CDir*> subtrees;
  mdcache->get_subtrees(subtrees);

  f->open_array_section("subtrees");
  for (const auto& dir : subtrees) {
    f->open_object_section("subtree");
    {
      f->dump_bool("is_auth", dir->is_auth());
      f->dump_int("auth_first", dir->get_dir_auth().first);
      f->dump_int("auth_second", dir->get_dir_auth().second);
      f->dump_int("export_pin", dir->inode->get_export_pin(false, false));
      f->dump_bool("distributed_ephemeral_pin", dir->inode->is_ephemeral_dist());
      f->dump_bool("random_ephemeral_pin", dir->inode->is_ephemeral_rand());
      f->dump_int("ephemeral_pin", mdcache->hash_into_rank_bucket(dir->inode->ino()));
      f->open_object_section("dir");
      dir->dump(f);
      f->close_section();
    }
    f->close_section();
  }
  f->close_section();
}


void MDSRank::command_export_dir(Formatter *f,
    std::string_view path,
    mds_rank_t target)
{
  int r = _command_export_dir(path, target);
  f->open_object_section("results");
  f->dump_int("return_code", r);
  f->close_section(); // results
}

int MDSRank::_command_export_dir(
    std::string_view path,
    mds_rank_t target)
{
  std::lock_guard l(mds_lock);
  filepath fp(path);

  if (target == whoami || !mdsmap->is_up(target) || !mdsmap->is_in(target)) {
    derr << "bad MDS target " << target << dendl;
    return -ENOENT;
  }

  CInode *in = mdcache->cache_traverse(fp);
  if (!in) {
    derr << "Bath path '" << path << "'" << dendl;
    return -ENOENT;
  }
  CDir *dir = in->get_dirfrag(frag_t());
  if (!dir || !(dir->is_auth())) {
    derr << "bad export_dir path dirfrag frag_t() or dir not auth" << dendl;
    return -EINVAL;
  }

  mdcache->migrator->export_dir(dir, target);
  return 0;
}

void MDSRank::command_dump_tree(const cmdmap_t &cmdmap, std::ostream &ss, Formatter *f) 
{
  std::string root;
  int64_t depth;
  cmd_getval(cmdmap, "root", root);
  if (!cmd_getval(cmdmap, "depth", depth))
    depth = -1;
  std::lock_guard l(mds_lock);
  CInode *in = mdcache->cache_traverse(filepath(root.c_str()));
  if (!in) {
    ss << "root inode is not in cache";
    return;
  }
  f->open_array_section("inodes");
  mdcache->dump_tree(in, 0, depth, f);
  f->close_section();
}

CDir *MDSRank::_command_dirfrag_get(
    const cmdmap_t &cmdmap,
    std::ostream &ss)
{
  std::string path;
  bool got = cmd_getval(cmdmap, "path", path);
  if (!got) {
    ss << "missing path argument";
    return NULL;
  }

  std::string frag_str;
  if (!cmd_getval(cmdmap, "frag", frag_str)) {
    ss << "missing frag argument";
    return NULL;
  }

  CInode *in = mdcache->cache_traverse(filepath(path.c_str()));
  if (!in) {
    // TODO really we should load something in if it's not in cache,
    // but the infrastructure is harder, and we might still be unable
    // to act on it if someone else is auth.
    ss << "directory '" << path << "' inode not in cache";
    return NULL;
  }

  frag_t fg;

  if (!fg.parse(frag_str.c_str())) {
    ss << "frag " << frag_str << " failed to parse";
    return NULL;
  }

  CDir *dir = in->get_dirfrag(fg);
  if (!dir) {
    ss << "frag " << in->ino() << "/" << fg << " not in cache ("
          "use `dirfrag ls` to see if it should exist)";
    return NULL;
  }

  if (!dir->is_auth()) {
    ss << "frag " << dir->dirfrag() << " not auth (auth = "
       << dir->authority() << ")";
    return NULL;
  }

  return dir;
}

bool MDSRank::command_dirfrag_split(
    cmdmap_t cmdmap,
    std::ostream &ss)
{
  std::lock_guard l(mds_lock);
  int64_t by = 0;
  if (!cmd_getval(cmdmap, "bits", by)) {
    ss << "missing bits argument";
    return false;
  }

  if (by <= 0) {
    ss << "must split by >0 bits";
    return false;
  }

  CDir *dir = _command_dirfrag_get(cmdmap, ss);
  if (!dir) {
    return false;
  }

  mdcache->split_dir(dir, by);

  return true;
}

bool MDSRank::command_dirfrag_merge(
    cmdmap_t cmdmap,
    std::ostream &ss)
{
  std::lock_guard l(mds_lock);
  std::string path;
  bool got = cmd_getval(cmdmap, "path", path);
  if (!got) {
    ss << "missing path argument";
    return false;
  }

  std::string frag_str;
  if (!cmd_getval(cmdmap, "frag", frag_str)) {
    ss << "missing frag argument";
    return false;
  }

  CInode *in = mdcache->cache_traverse(filepath(path.c_str()));
  if (!in) {
    ss << "directory '" << path << "' inode not in cache";
    return false;
  }

  frag_t fg;
  if (!fg.parse(frag_str.c_str())) {
    ss << "frag " << frag_str << " failed to parse";
    return false;
  }

  mdcache->merge_dir(in, fg);

  return true;
}

bool MDSRank::command_dirfrag_ls(
    cmdmap_t cmdmap,
    std::ostream &ss,
    Formatter *f)
{
  std::lock_guard l(mds_lock);
  std::string path;
  bool got = cmd_getval(cmdmap, "path", path);
  if (!got) {
    ss << "missing path argument";
    return false;
  }

  CInode *in = mdcache->cache_traverse(filepath(path.c_str()));
  if (!in) {
    ss << "directory inode not in cache";
    return false;
  }

  f->open_array_section("frags");
  frag_vec_t leaves;
  // NB using get_leaves_under instead of get_dirfrags to give
  // you the list of what dirfrags may exist, not which are in cache
  in->dirfragtree.get_leaves_under(frag_t(), leaves);
  for (const auto& leaf : leaves) {
    f->open_object_section("frag");
    f->dump_int("value", leaf.value());
    f->dump_int("bits", leaf.bits());
    CachedStackStringStream css;
    *css << std::hex << leaf.value() << "/" << std::dec << leaf.bits();
    f->dump_string("str", css->strv());
    f->close_section();
  }
  f->close_section();

  return true;
}

void MDSRank::command_openfiles_ls(Formatter *f) 
{
  std::lock_guard l(mds_lock);
  mdcache->dump_openfiles(f);
}

void MDSRank::command_dump_inode(Formatter *f, const cmdmap_t &cmdmap, std::ostream &ss)
{
  std::lock_guard l(mds_lock);
  int64_t number;
  bool got = cmd_getval(cmdmap, "number", number);
  if (!got) {
    ss << "missing inode number";
    return;
  }
  
  bool success = mdcache->dump_inode(f, number);
  if (!success) {
    ss << "dump inode failed, wrong inode number or the inode is not cached";
  }
}

void MDSRank::dump_status(Formatter *f) const
{
  if (state == MDSMap::STATE_REPLAY ||
      state == MDSMap::STATE_STANDBY_REPLAY) {
    mdlog->dump_replay_status(f);
  } else if (state == MDSMap::STATE_RESOLVE) {
    mdcache->dump_resolve_status(f);
  } else if (state == MDSMap::STATE_RECONNECT) {
    server->dump_reconnect_status(f);
  } else if (state == MDSMap::STATE_REJOIN) {
    mdcache->dump_rejoin_status(f);
  } else if (state == MDSMap::STATE_CLIENTREPLAY) {
    dump_clientreplay_status(f);
  }
  f->dump_float("rank_uptime", get_uptime().count());
}

void MDSRank::dump_clientreplay_status(Formatter *f) const
{
  f->open_object_section("clientreplay_status");
  f->dump_unsigned("clientreplay_queue", replay_queue.size());
  f->dump_unsigned("active_replay", mdcache->get_num_client_requests());
  f->close_section();
}

void MDSRankDispatcher::update_log_config()
{
  map<string,string> log_to_monitors;
  map<string,string> log_to_syslog;
  map<string,string> log_channel;
  map<string,string> log_prio;
  map<string,string> log_to_graylog;
  map<string,string> log_to_graylog_host;
  map<string,string> log_to_graylog_port;
  uuid_d fsid;
  string host;

  if (parse_log_client_options(g_ceph_context, log_to_monitors, log_to_syslog,
                               log_channel, log_prio, log_to_graylog,
                               log_to_graylog_host, log_to_graylog_port,
                               fsid, host) == 0)
    clog->update_config(log_to_monitors, log_to_syslog,
                        log_channel, log_prio, log_to_graylog,
                        log_to_graylog_host, log_to_graylog_port,
                        fsid, host);
  dout(10) << __func__ << " log_to_monitors " << log_to_monitors << dendl;
}

void MDSRank::create_logger()
{
  dout(10) << "create_logger" << dendl;
  {
    PerfCountersBuilder mds_plb(g_ceph_context, "mds", l_mds_first, l_mds_last);

    // super useful (high prio) perf stats
    mds_plb.add_u64_counter(l_mds_request, "request", "Requests", "req",
                            PerfCountersBuilder::PRIO_CRITICAL);
    mds_plb.add_time_avg(l_mds_reply_latency, "reply_latency", "Reply latency", "rlat",
                         PerfCountersBuilder::PRIO_CRITICAL);
    mds_plb.add_u64(l_mds_inodes, "inodes", "Inodes", "inos",
                    PerfCountersBuilder::PRIO_CRITICAL);
    mds_plb.add_u64_counter(l_mds_forward, "forward", "Forwarding request", "fwd",
                            PerfCountersBuilder::PRIO_INTERESTING);
    mds_plb.add_u64(l_mds_caps, "caps", "Capabilities", "caps",
                    PerfCountersBuilder::PRIO_INTERESTING);
    mds_plb.add_u64_counter(l_mds_exported_inodes, "exported_inodes", "Exported inodes",
                            "exi", PerfCountersBuilder::PRIO_INTERESTING);
    mds_plb.add_u64_counter(l_mds_imported_inodes, "imported_inodes", "Imported inodes",
                            "imi", PerfCountersBuilder::PRIO_INTERESTING);

    // useful dir/inode/subtree stats
    mds_plb.set_prio_default(PerfCountersBuilder::PRIO_USEFUL);
    mds_plb.add_u64(l_mds_root_rfiles, "root_rfiles", "root inode rfiles");
    mds_plb.add_u64(l_mds_root_rbytes, "root_rbytes", "root inode rbytes");
    mds_plb.add_u64(l_mds_root_rsnaps, "root_rsnaps", "root inode rsnaps");
    mds_plb.add_u64_counter(l_mds_dir_fetch, "dir_fetch", "Directory fetch");
    mds_plb.add_u64_counter(l_mds_dir_commit, "dir_commit", "Directory commit");
    mds_plb.add_u64_counter(l_mds_dir_split, "dir_split", "Directory split");
    mds_plb.add_u64_counter(l_mds_dir_merge, "dir_merge", "Directory merge");
    mds_plb.add_u64(l_mds_inodes_pinned, "inodes_pinned", "Inodes pinned");
    mds_plb.add_u64(l_mds_inodes_expired, "inodes_expired", "Inodes expired");
    mds_plb.add_u64(l_mds_inodes_with_caps, "inodes_with_caps",
                    "Inodes with capabilities");
    mds_plb.add_u64(l_mds_subtrees, "subtrees", "Subtrees");
    mds_plb.add_u64(l_mds_load_cent, "load_cent", "Load per cent");
    mds_plb.add_u64_counter(l_mds_openino_dir_fetch, "openino_dir_fetch",
                            "OpenIno incomplete directory fetchings");

    // low prio stats
    mds_plb.set_prio_default(PerfCountersBuilder::PRIO_DEBUGONLY);
    mds_plb.add_u64_counter(l_mds_reply, "reply", "Replies");
    mds_plb.add_u64(l_mds_inodes_top, "inodes_top", "Inodes on top");
    mds_plb.add_u64(l_mds_inodes_bottom, "inodes_bottom", "Inodes on bottom");
    mds_plb.add_u64(
      l_mds_inodes_pin_tail, "inodes_pin_tail", "Inodes on pin tail");
    mds_plb.add_u64_counter(l_mds_traverse, "traverse", "Traverses");
    mds_plb.add_u64_counter(l_mds_traverse_hit, "traverse_hit", "Traverse hits");
    mds_plb.add_u64_counter(l_mds_traverse_forward, "traverse_forward",
                            "Traverse forwards");
    mds_plb.add_u64_counter(l_mds_traverse_discover, "traverse_discover",
                            "Traverse directory discovers");
    mds_plb.add_u64_counter(l_mds_traverse_dir_fetch, "traverse_dir_fetch",
                            "Traverse incomplete directory content fetchings");
    mds_plb.add_u64_counter(l_mds_traverse_remote_ino, "traverse_remote_ino",
                            "Traverse remote dentries");
    mds_plb.add_u64_counter(l_mds_traverse_lock, "traverse_lock",
                            "Traverse locks");
    mds_plb.add_u64(l_mds_dispatch_queue_len, "q", "Dispatch queue length");
    mds_plb.add_u64_counter(l_mds_exported, "exported", "Exports");
    mds_plb.add_u64_counter(l_mds_imported, "imported", "Imports");
    mds_plb.add_u64_counter(l_mds_openino_backtrace_fetch, "openino_backtrace_fetch",
                            "OpenIno backtrace fetchings");
    mds_plb.add_u64_counter(l_mds_openino_peer_discover, "openino_peer_discover",
                            "OpenIno peer inode discovers");

    logger = mds_plb.create_perf_counters();
    g_ceph_context->get_perfcounters_collection()->add(logger);
  }

  {
    PerfCountersBuilder mdm_plb(g_ceph_context, "mds_mem", l_mdm_first, l_mdm_last);
    mdm_plb.add_u64(l_mdm_ino, "ino", "Inodes", "ino",
                    PerfCountersBuilder::PRIO_INTERESTING);
    mdm_plb.add_u64(l_mdm_dn, "dn", "Dentries", "dn",
                    PerfCountersBuilder::PRIO_INTERESTING);

    mdm_plb.set_prio_default(PerfCountersBuilder::PRIO_USEFUL);
    mdm_plb.add_u64_counter(l_mdm_inoa, "ino+", "Inodes opened");
    mdm_plb.add_u64_counter(l_mdm_inos, "ino-", "Inodes closed");
    mdm_plb.add_u64(l_mdm_dir, "dir", "Directories");
    mdm_plb.add_u64_counter(l_mdm_dira, "dir+", "Directories opened");
    mdm_plb.add_u64_counter(l_mdm_dirs, "dir-", "Directories closed");
    mdm_plb.add_u64_counter(l_mdm_dna, "dn+", "Dentries opened");
    mdm_plb.add_u64_counter(l_mdm_dns, "dn-", "Dentries closed");
    mdm_plb.add_u64(l_mdm_cap, "cap", "Capabilities");
    mdm_plb.add_u64_counter(l_mdm_capa, "cap+", "Capabilities added");
    mdm_plb.add_u64_counter(l_mdm_caps, "cap-", "Capabilities removed");
    mdm_plb.add_u64(l_mdm_heap, "heap", "Heap size");

    mdm_plb.set_prio_default(PerfCountersBuilder::PRIO_DEBUGONLY);
    mdm_plb.add_u64(l_mdm_rss, "rss", "RSS");

    mlogger = mdm_plb.create_perf_counters();
    g_ceph_context->get_perfcounters_collection()->add(mlogger);
  }

  mdlog->create_logger();
  server->create_logger();
  purge_queue.create_logger();
  sessionmap.register_perfcounters();
  mdcache->register_perfcounters();
}

void MDSRank::check_ops_in_flight()
{
  string summary;
  vector<string> warnings;
  int slow = 0;
  if (op_tracker.check_ops_in_flight(&summary, warnings, &slow)) {
    clog->warn() << summary;
    for (const auto& warning : warnings) {
      clog->warn() << warning;
    }
  }
 
  // set mds slow request count 
  mds_slow_req_count = slow;
  return;
}

void MDSRankDispatcher::handle_osd_map()
{
  if (is_active() &&
      mdsmap->get_tableserver() == whoami) {
    snapserver->check_osd_map(true);
  }

  server->handle_osd_map();

  purge_queue.update_op_limit(*mdsmap);

  std::set<entity_addr_t> newly_blacklisted;
  objecter->consume_blacklist_events(&newly_blacklisted);
  auto epoch = objecter->with_osdmap([](const OSDMap &o){return o.get_epoch();});
  dout(4) << "handle_osd_map epoch " << epoch << ", "
          << newly_blacklisted.size() << " new blacklist entries" << dendl;
  auto victims = server->apply_blacklist(newly_blacklisted);
  if (victims) {
    set_osd_epoch_barrier(epoch);
  }


  // By default the objecter only requests OSDMap updates on use,
  // we would like to always receive the latest maps in order to
  // apply policy based on the FULL flag.
  objecter->maybe_request_map();
}

int MDSRank::config_client(int64_t session_id, bool remove,
                           const std::string& option, const std::string& value,
                           std::ostream& ss)
{
  Session *session = sessionmap.get_session(entity_name_t(CEPH_ENTITY_TYPE_CLIENT, session_id));
  if (!session) {
    ss << "session " << session_id << " not in sessionmap!";
    return -ENOENT;
  }

  if (option == "timeout") {
    if (remove) {
      auto it = session->info.client_metadata.find("timeout");
      if (it == session->info.client_metadata.end()) {
        ss << "Nonexistent config: " << option;
        return -ENODATA;
      }
      session->info.client_metadata.erase(it);
    } else {
      char *end;
      strtoul(value.c_str(), &end, 0);
      if (*end) {
        ss << "Invalid config for timeout: " << value;
        return -EINVAL;
      }
      session->info.client_metadata[option] = value;
    }
    //sessionmap._mark_dirty(session, true);
  } else {
    ss << "Invalid config option: " << option;
    return -EINVAL;
  }

  return 0;
}

bool MDSRank::evict_client(int64_t session_id,
    bool wait, bool blacklist, std::ostream& err_ss,
    Context *on_killed)
{
  ceph_assert(ceph_mutex_is_locked_by_me(mds_lock));

  // Mutually exclusive args
  ceph_assert(!(wait && on_killed != nullptr));

  if (is_any_replay()) {
    err_ss << "MDS is replaying log";
    return false;
  }

  Session *session = sessionmap.get_session(
      entity_name_t(CEPH_ENTITY_TYPE_CLIENT, session_id));
  if (!session) {
    err_ss << "session " << session_id << " not in sessionmap!";
    return false;
  }

  auto& addr = session->info.inst.addr;
  {
    CachedStackStringStream css;
    *css << "Evicting " << (blacklist ? "(and blacklisting) " : "")
         << "client session " << session_id << " (" << addr << ")";
    dout(1) << css->strv() << dendl;
    clog->info() << css->strv();
  }

  dout(4) << "Preparing blacklist command... (wait=" << wait << ")" << dendl;
  stringstream ss;
  ss << "{\"prefix\":\"osd blacklist\", \"blacklistop\":\"add\",";
  ss << "\"addr\":\"";
  ss << addr;
  ss << "\"}";
  std::string tmp = ss.str();
  std::vector<std::string> cmd = {tmp};

  auto kill_client_session = [this, session_id, wait, on_killed](){
    ceph_assert(ceph_mutex_is_locked_by_me(mds_lock));
    Session *session = sessionmap.get_session(
        entity_name_t(CEPH_ENTITY_TYPE_CLIENT, session_id));
    if (session) {
      if (on_killed || !wait) {
        server->kill_session(session, on_killed);
      } else {
        C_SaferCond on_safe;
        server->kill_session(session, &on_safe);

        mds_lock.unlock();
        on_safe.wait();
        mds_lock.lock();
      }
    } else {
      dout(1) << "session " << session_id << " was removed while we waited "
      "for blacklist" << dendl;

      // Even though it wasn't us that removed it, kick our completion
      // as the session has been removed.
      if (on_killed) {
        on_killed->complete(0);
      }
    }
  };

  auto apply_blacklist = [this, cmd](std::function<void ()> fn){
    ceph_assert(ceph_mutex_is_locked_by_me(mds_lock));

    Context *on_blacklist_done = new LambdaContext([this, fn](int r) {
      objecter->wait_for_latest_osdmap(
       new C_OnFinisher(
         new LambdaContext([this, fn](int r) {
              std::lock_guard l(mds_lock);
              auto epoch = objecter->with_osdmap([](const OSDMap &o){
                  return o.get_epoch();
              });

              set_osd_epoch_barrier(epoch);

              fn();
            }), finisher)
       );
    });

    dout(4) << "Sending mon blacklist command: " << cmd[0] << dendl;
    monc->start_mon_command(cmd, {}, nullptr, nullptr, on_blacklist_done);
  };

  if (wait) {
    if (blacklist) {
      C_SaferCond inline_ctx;
      apply_blacklist([&inline_ctx](){inline_ctx.complete(0);});
      mds_lock.unlock();
      inline_ctx.wait();
      mds_lock.lock();
    }

    // We dropped mds_lock, so check that session still exists
    session = sessionmap.get_session(entity_name_t(CEPH_ENTITY_TYPE_CLIENT,
          session_id));
    if (!session) {
      dout(1) << "session " << session_id << " was removed while we waited "
                 "for blacklist" << dendl;
      return true;
    }
    kill_client_session();
  } else {
    if (blacklist) {
      apply_blacklist(kill_client_session);
    } else {
      kill_client_session();
    }
  }

  return true;
}

void MDSRank::bcast_mds_map()
{
  dout(7) << "bcast_mds_map " << mdsmap->get_epoch() << dendl;

  // share the map with mounted clients
  set<Session*> clients;
  sessionmap.get_client_session_set(clients);
  for (const auto &session : clients) {
    auto m = make_message<MMDSMap>(monc->get_fsid(), *mdsmap);
    session->get_connection()->send_message2(std::move(m));
  }
  last_client_mdsmap_bcast = mdsmap->get_epoch();
}

MDSRankDispatcher::MDSRankDispatcher(
    mds_rank_t whoami_,
    ceph::mutex &mds_lock_,
    LogChannelRef &clog_,
    SafeTimer &timer_,
    Beacon &beacon_,
    std::unique_ptr<MDSMap> &mdsmap_,
    Messenger *msgr,
    MonClient *monc_,
    MgrClient *mgrc,
    Context *respawn_hook_,
    Context *suicide_hook_)
  : MDSRank(whoami_, mds_lock_, clog_, timer_, beacon_, mdsmap_,
            msgr, monc_, mgrc, respawn_hook_, suicide_hook_)
{
    g_conf().add_observer(this);
}

void MDSRank::command_cache_drop(uint64_t timeout, Formatter *f, Context *on_finish) {
  dout(20) << __func__ << dendl;

  std::lock_guard locker(mds_lock);
  C_Drop_Cache *request = new C_Drop_Cache(server, mdcache, mdlog, this,
                                           timeout, f, on_finish);
  request->send();
}

epoch_t MDSRank::get_osd_epoch() const
{
  return objecter->with_osdmap(std::mem_fn(&OSDMap::get_epoch));  
}

const char** MDSRankDispatcher::get_tracked_conf_keys() const
{
  static const char* KEYS[] = {
    "clog_to_graylog",
    "clog_to_graylog_host",
    "clog_to_graylog_port",
    "clog_to_monitors",
    "clog_to_syslog",
    "clog_to_syslog_facility",
    "clog_to_syslog_level",
    "fsid",
    "host",
    "mds_bal_fragment_dirs",
    "mds_bal_fragment_interval",
    "mds_cache_memory_limit",
    "mds_cache_mid",
    "mds_cache_reservation",
    "mds_cache_trim_decay_rate",
    "mds_cap_revoke_eviction_timeout",
    "mds_dump_cache_threshold_file",
    "mds_dump_cache_threshold_formatter",
    "mds_enable_op_tracker",
    "mds_export_ephemeral_random",
    "mds_export_ephemeral_random_max",
    "mds_export_ephemeral_distributed",
    "mds_health_cache_threshold",
    "mds_inject_migrator_session_race",
    "mds_log_pause",
    "mds_max_export_size",
    "mds_max_purge_files",
    "mds_forward_all_requests_to_auth",
    "mds_max_purge_ops",
    "mds_max_purge_ops_per_pg",
    "mds_max_snaps_per_dir",
    "mds_op_complaint_time",
    "mds_op_history_duration",
    "mds_op_history_size",
    "mds_op_log_threshold",
    "mds_recall_max_decay_rate",
    "mds_recall_warning_decay_rate",
    "mds_request_load_average_decay_rate",
    "mds_session_cache_liveness_decay_rate",
    "mds_replay_unsafe_with_closed_session",
    NULL
  };
  return KEYS;
}

void MDSRankDispatcher::handle_conf_change(const ConfigProxy& conf, const std::set<std::string>& changed)
{
  // XXX with or without mds_lock!

  if (changed.count("mds_op_complaint_time") || changed.count("mds_op_log_threshold")) {
    op_tracker.set_complaint_and_threshold(conf->mds_op_complaint_time, conf->mds_op_log_threshold);
  }
  if (changed.count("mds_op_history_size") || changed.count("mds_op_history_duration")) {
    op_tracker.set_history_size_and_duration(conf->mds_op_history_size, conf->mds_op_history_duration);
  }
  if (changed.count("mds_enable_op_tracker")) {
    op_tracker.set_tracking(conf->mds_enable_op_tracker);
  }
  if (changed.count("clog_to_monitors") ||
      changed.count("clog_to_syslog") ||
      changed.count("clog_to_syslog_level") ||
      changed.count("clog_to_syslog_facility") ||
      changed.count("clog_to_graylog") ||
      changed.count("clog_to_graylog_host") ||
      changed.count("clog_to_graylog_port") ||
      changed.count("host") ||
      changed.count("fsid")) {
    update_log_config();
  }

  finisher->queue(new LambdaContext([this, changed](int) {
    std::scoped_lock lock(mds_lock);

    dout(10) << "flushing conf change to components: " << changed << dendl;

    if (changed.count("mds_log_pause") && !g_conf()->mds_log_pause) {
      mdlog->kick_submitter();
    }
    sessionmap.handle_conf_change(changed);
    server->handle_conf_change(changed);
    mdcache->handle_conf_change(changed, *mdsmap);
    purge_queue.handle_conf_change(changed, *mdsmap);
  }));
}

void MDSRank::get_task_status(std::map<std::string, std::string> *status) {
  dout(20) << __func__ << dendl;

  // scrub summary for now..
  std::string_view scrub_summary = scrubstack->scrub_summary();
  status->emplace(SCRUB_STATUS_KEY, std::move(scrub_summary));
}

void MDSRank::schedule_update_timer_task() {
  dout(20) << __func__ << dendl;

  timer.add_event_after(g_conf().get_val<double>("mds_task_status_update_interval"),
                        new LambdaContext([this](int) {
                            send_task_status();
                          }));
}

void MDSRank::send_task_status() {
  std::map<std::string, std::string> status;
  get_task_status(&status);

  if (!status.empty()) {
    dout(20) << __func__ << ": updating " << status.size() << " status keys" << dendl;

    int r = mgrc->service_daemon_update_task_status(std::move(status));
    if (r < 0) {
      derr << ": failed to update service daemon status: " << cpp_strerror(r) << dendl;
    }
  }

  schedule_update_timer_task();
}