diskless replication on slave side (don't store rdb to file), plus some other related fixes

The implementation of the diskless replication was currently diskless only on the master side.
The slave side was still storing the received rdb file to the disk before loading it back in and parsing it.

This commit adds two modes to load rdb directly from socket:
1) when-empty
2) using "swapdb"
the third mode of using diskless slave by flushdb is risky and currently not included.

other changes:
--------------
distinguish between aof configuration and state so that we can re-enable aof only when sync eventually
succeeds (and not when exiting from readSyncBulkPayload after a failed attempt)
also a CONFIG GET and INFO during rdb loading would have lied

When loading rdb from the network, don't kill the server on short read (that can be a network error)

Fix rdb check when performed on preamble AOF

tests:
run replication tests for diskless slave too
make replication test a bit more aggressive
Add test for diskless load swapdb

src/rdb.c

@@ -44,6 +44,7 @@
 
 #define rdbExitReportCorruptRDB(...) rdbCheckThenExit(__LINE__,__VA_ARGS__)
 
+char* rdbFileBeingLoaded = NULL; /* used for rdb checking on read error */
 extern int rdbCheckMode;
 void rdbCheckError(const char *fmt, ...);
 void rdbCheckSetError(const char *fmt, ...);
@@ -61,11 +62,17 @@ void rdbCheckThenExit(int linenum, char *reason, ...) {
 
     if (!rdbCheckMode) {
         serverLog(LL_WARNING, "%s", msg);
-        char *argv[2] = {"",server.rdb_filename};
-        redis_check_rdb_main(2,argv,NULL);
+        if (rdbFileBeingLoaded) {
+            char *argv[2] = {"",rdbFileBeingLoaded};
+            redis_check_rdb_main(2,argv,NULL);
+        } else {
+            serverLog(LL_WARNING, "Failure loading rdb format from socket, assuming connection error, resuming operation.");
+            return;
+        }
     } else {
         rdbCheckError("%s",msg);
     }
+    serverLog(LL_WARNING, "Terminating server after rdb file reading failure.");
     exit(1);
 }
 
@@ -1039,6 +1046,11 @@ int rdbSaveKeyValuePair(rio *rdb, robj *key, robj *val, long long expiretime) {
     if (rdbSaveObjectType(rdb,val) == -1) return -1;
     if (rdbSaveStringObject(rdb,key) == -1) return -1;
     if (rdbSaveObject(rdb,val,key) == -1) return -1;
+
+    /* Delay return if required (for testing) */
+    if (server.rdb_key_save_delay)
+        usleep(server.rdb_key_save_delay);
+
     return 1;
 }
 
@@ -1800,18 +1812,23 @@ robj *rdbLoadObject(int rdbtype, rio *rdb, robj *key) {
 
 /* Mark that we are loading in the global state and setup the fields
  * needed to provide loading stats. */
-void startLoading(FILE *fp) {
-    struct stat sb;
-
+void startLoading(size_t size) {
     /* Load the DB */
     server.loading = 1;
     server.loading_start_time = time(NULL);
     server.loading_loaded_bytes = 0;
-    if (fstat(fileno(fp), &sb) == -1) {
-        server.loading_total_bytes = 0;
-    } else {
-        server.loading_total_bytes = sb.st_size;
-    }
+    server.loading_total_bytes = size;
+}
+
+/* Mark that we are loading in the global state and setup the fields
+ * needed to provide loading stats.
+ * 'filename' is optional and used for rdb-check on error */
+void startLoadingFile(FILE *fp, char* filename) {
+    struct stat sb;
+    if (fstat(fileno(fp), &sb) == -1)
+        sb.st_size = 0;
+    rdbFileBeingLoaded = filename;
+    startLoading(sb.st_size);
 }
 
 /* Refresh the loading progress info */
@@ -1824,6 +1841,7 @@ void loadingProgress(off_t pos) {
 /* Loading finished */
 void stopLoading(void) {
     server.loading = 0;
+    rdbFileBeingLoaded = NULL;
 }
 
 /* Track loading progress in order to serve client's from time to time
@@ -2089,7 +2107,7 @@ int rdbLoad(char *filename, rdbSaveInfo *rsi) {
     int retval;
 
     if ((fp = fopen(filename,"r")) == NULL) return C_ERR;
-    startLoading(fp);
+    startLoadingFile(fp, filename);
     rioInitWithFile(&rdb,fp);
     retval = rdbLoadRio(&rdb,rsi,0);
     fclose(fp);