From b541ccef25d20749c739b858ef8694512b7f3bce Mon Sep 17 00:00:00 2001
From: "zhaozhao.zz" <zhaozhao.zz@alibaba-inc.com>
Date: Wed, 20 Sep 2017 13:47:42 +0800
Subject: [PATCH] PSYNC2: make persisiting replication info more solid

This commit is a reinforcement of commit c1c99e9.

1. Replication information can be stored when the RDB file is
generated by a mater using server.slaveseldb when server.repl_backlog
is not NULL, or set repl_stream_db be -1. That's safe, because
NULL server.repl_backlog will trigger full synchronization,
then master will send SELECT command to replicaiton stream.
2. Only do rdbSave* when rsiptr is not NULL,
if we do rdbSave* without rdbSaveInfo, slave will miss repl-stream-db.
3. Save the replication informations also in the case of
SAVE command, FLUSHALL command and DEBUG reload.
---
 src/db.c          |  4 +++-
 src/debug.c       |  4 +++-
 src/rdb.c         | 20 +++++++++++++++++---
 src/replication.c | 14 ++++++++++----
 4 files changed, 33 insertions(+), 9 deletions(-)

diff --git a/src/db.c b/src/db.c
index 7d1504d3..71c642d0 100644
--- a/src/db.c
+++ b/src/db.c
@@ -416,7 +416,9 @@ void flushallCommand(client *c) {
         /* Normally rdbSave() will reset dirty, but we don't want this here
          * as otherwise FLUSHALL will not be replicated nor put into the AOF. */
         int saved_dirty = server.dirty;
-        rdbSave(server.rdb_filename,NULL);
+        rdbSaveInfo rsi, *rsiptr;
+        rsiptr = rdbPopulateSaveInfo(&rsi);
+        rdbSave(server.rdb_filename,rsiptr);
         server.dirty = saved_dirty;
     }
     server.dirty++;
diff --git a/src/debug.c b/src/debug.c
index d6e12ec2..5c3fd347 100644
--- a/src/debug.c
+++ b/src/debug.c
@@ -335,7 +335,9 @@ void debugCommand(client *c) {
         if (c->argc >= 3) c->argv[2] = tryObjectEncoding(c->argv[2]);
         serverAssertWithInfo(c,c->argv[0],1 == 2);
     } else if (!strcasecmp(c->argv[1]->ptr,"reload")) {
-        if (rdbSave(server.rdb_filename,NULL) != C_OK) {
+        rdbSaveInfo rsi, *rsiptr;
+        rsiptr = rdbPopulateSaveInfo(&rsi);
+        if (rdbSave(server.rdb_filename,rsiptr) != C_OK) {
             addReply(c,shared.err);
             return;
         }
diff --git a/src/rdb.c b/src/rdb.c
index 9c2acf7b..c83a05a3 100644
--- a/src/rdb.c
+++ b/src/rdb.c
@@ -1975,7 +1975,9 @@ void saveCommand(client *c) {
         addReplyError(c,"Background save already in progress");
         return;
     }
-    if (rdbSave(server.rdb_filename,NULL) == C_OK) {
+    rdbSaveInfo rsi, *rsiptr;
+    rsiptr = rdbPopulateSaveInfo(&rsi);
+    if (rdbSave(server.rdb_filename,rsiptr) == C_OK) {
         addReply(c,shared.ok);
     } else {
         addReply(c,shared.err);
@@ -2017,21 +2019,33 @@ void bgsaveCommand(client *c) {
 }
 
 /* Populate the rdbSaveInfo structure used to persist the replication
- * information inside the RDB file. Currently the structure explicitly
+ * information inside the RDB file.
+ * For master, if server.repl_backlog is not NULL, fill rdbSaveInfo with
+ * server.slaveseldb, otherwise just use init -1.
+ * Don't worry, master will send SELECT command to replication stream,
+ * because if server.repl_backlog is NULL, that will trigger full synchronization,
+ * function replicationSetupSlaveForFullResync() sets server.slaveseldb be -1,
+ * then replicationFeedSlaves() will send SELECT command when server.slaveseldb is -1.
+ * For slave, currently the structure explicitly
  * contains just the currently selected DB from the master stream, however
  * if the rdbSave*() family functions receive a NULL rsi structure also
  * the Replication ID/offset is not saved. The function popultes 'rsi'
  * that is normally stack-allocated in the caller, returns the populated
  * pointer if the instance has a valid master client, otherwise NULL
- * is returned, and the RDB savign wil not persist any replication related
+ * is returned, and the RDB saving will not persist any replication related
  * information. */
 rdbSaveInfo *rdbPopulateSaveInfo(rdbSaveInfo *rsi) {
     rdbSaveInfo rsi_init = RDB_SAVE_INFO_INIT;
     *rsi = rsi_init;
+    if (!server.masterhost) {
+        if (server.repl_backlog) rsi->repl_stream_db = server.slaveseldb;
+        return rsi;
+    }
     if (server.master) {
         rsi->repl_stream_db = server.master->db->id;
         return rsi;
     } else {
+        serverLog(LL_WARNING,"As a slave there is no valid master, can not persist replication information");
         return NULL;
     }
 }
diff --git a/src/replication.c b/src/replication.c
index 2b2673a4..4358d63e 100644
--- a/src/replication.c
+++ b/src/replication.c
@@ -571,10 +571,16 @@ int startBgsaveForReplication(int mincapa) {
 
     rdbSaveInfo rsi, *rsiptr;
     rsiptr = rdbPopulateSaveInfo(&rsi);
-    if (socket_target)
-        retval = rdbSaveToSlavesSockets(rsiptr);
-    else
-        retval = rdbSaveBackground(server.rdb_filename,rsiptr);
+    /* Only do rdbSave* when rsiptr is not NULL,
+     * otherwise slave will miss repl-stream-db. */
+    if (rsiptr) {
+        if (socket_target)
+            retval = rdbSaveToSlavesSockets(rsiptr);
+        else
+            retval = rdbSaveBackground(server.rdb_filename,rsiptr);
+    } else {
+        retval = C_ERR;
+    }
 
     /* If we failed to BGSAVE, remove the slaves waiting for a full
      * resynchorinization from the list of salves, inform them with