mon/Paxos: break commit() into two pieces

One part happens before the txn starts, the other after.  Move all of the
internal state update to the bottom half.  Eventually this will matter.

Signed-off-by: Sage Weil <sage@redhat.com>

diff --git a/src/mon/Paxos.cc b/src/mon/Paxos.cc
index 4635c0ad0bfa3810a77868f6412bef299def7a5a..f3ddb5ed8467d1bc7411473f7b5e7a077e0062f8 100644 (file)
--- a/src/mon/Paxos.cc
+++ b/src/mon/Paxos.cc
@@ -665,7 +665,7 @@ void Paxos::begin(bufferlist& v)
 
   if (mon->get_quorum().size() == 1) {
     // we're alone, take it easy
-    commit();
+    commit_start();
     if (do_refresh()) {
       assert(is_updating());  // we can't be updating-previous with quorum of 1
       commit_proposal();
@@ -793,7 +793,7 @@ void Paxos::handle_accept(MMonPaxos *accept)
   if (accepted == mon->get_quorum()) {
     // yay, commit!
     dout(10) << " got majority, committing, done with update" << dendl;
-    commit();
+    commit_start();
     if (!do_refresh())
       goto out;
     if (is_updating())
@@ -831,23 +831,16 @@ void Paxos::accept_timeout()
   mon->bootstrap();
 }
 
-void Paxos::commit()
+void Paxos::commit_start()
 {
-  dout(10) << "commit " << last_committed+1 << dendl;
-
-  // cancel lease - it was for the old value.
-  //  (this would only happen if message layer lost the 'begin', but
-  //   leader still got a majority and committed with out us.)
-  lease_expire = utime_t();  // cancel lease
+  dout(10) << __func__ << " " << (last_committed+1) << dendl;
 
   assert(g_conf->paxos_kill_at != 7);
 
   MonitorDBStore::TransactionRef t(new MonitorDBStore::Transaction);
 
   // commit locally
-  last_committed++;
-  last_commit_time = ceph_clock_now(g_ceph_context);
-  t->put(get_name(), "last_committed", last_committed);
+  t->put(get_name(), "last_committed", last_committed + 1);
 
   // decode the value and apply its transaction to the store.
   // this value can now be read from last_committed.
@@ -862,15 +855,29 @@ void Paxos::commit()
   logger->inc(l_paxos_commit);
   logger->inc(l_paxos_commit_keys, t->get_keys());
   logger->inc(l_paxos_commit_bytes, t->get_bytes());
-  utime_t start = ceph_clock_now(NULL);
+  commit_start_stamp = ceph_clock_now(NULL);
 
   get_store()->apply_transaction(t);
 
+  commit_finish();
+}
+
+void Paxos::commit_finish()
+{
+  dout(20) << __func__ << " " << (last_committed+1) << dendl;
   utime_t end = ceph_clock_now(NULL);
-  logger->tinc(l_paxos_commit_latency, end - start);
+  logger->tinc(l_paxos_commit_latency, end - commit_start_stamp);
 
   assert(g_conf->paxos_kill_at != 8);
 
+  // cancel lease - it was for the old value.
+  //  (this would only happen if message layer lost the 'begin', but
+  //   leader still got a majority and committed with out us.)
+  lease_expire = utime_t();  // cancel lease
+
+  last_committed++;
+  last_commit_time = ceph_clock_now(NULL);
+
   // refresh first_committed; this txn may have trimmed.
   first_committed = get_store()->get(get_name(), "first_committed");
 

diff --git a/src/mon/Paxos.h b/src/mon/Paxos.h
index c38debd9fdcd8c95096c682d449a60f7473452a2..ed345a4144f811818bd32af49d01440f69935205 100644 (file)
--- a/src/mon/Paxos.h
+++ b/src/mon/Paxos.h
@@ -320,7 +320,7 @@ private:
   /**
    * Last committed value's time.
    *
-   * When the commit happened.
+   * When the commit finished.
    */
   utime_t last_commit_time;
   /**
@@ -870,6 +870,9 @@ private:
    * @}
    */
 
+
+  utime_t commit_start_stamp;
+
   /**
    * Commit a value throughout the system.
    *
@@ -883,7 +886,8 @@ private:
    * @post Value locally stored
    * @post Quorum members instructed to commit the new value.
    */
-  void commit();
+  void commit_start();
+  void commit_finish();   ///< finish a commit after txn becomes durable
   /**
    * Commit the new value to stable storage as being the latest available
    * version.